REFLECTION AND TRANSMISSION OF OBLIQUELY INCIDENT LIGHT BY ASYMMETRIC SERIAL

The Pennsylvania State University

The Graduate School

College of Engineering

REFLECTION AND TRANSMISSION OF OBLIQUELY INCIDENT

LIGHT BY ASYMMETRIC SERIAL-BIDEPOSITED CHIRAL

SCULPTURED THIN FILMS

A Thesis in

Engineering Science and Mechanics

by

Patrick David McAtee

© 2016 Patrick David McAtee

Submitted in Partial Fulfillment

of the Requirements

for the Degree of

Master of Science

May 2016

The thesis of Patrick David McAtee was reviewed and approved∗ by the following:

Akhlesh Lakhtakia

Charles Godfrey Binder Professor of Engineering Science and Mechanics

Thesis Co-Advisor

Jian Xu

Associate Professor of Engineering Science and Mechanics and Adjunct

Professor of Electrical Engineering

Thesis Co-Advisor

Mark W. Horn

Professor of Engineering Science and Mechanics

Judith A. Todd

P. B. Breneman Professor and Head of the Department of Engineering

Science and Mechanics

∗Signatures are on file in the Graduate School.

ii

Abstract

A chiral sculptured thin film (STF) exhibits the Circular Bragg Phenomenon(CBP), which is the differential reflection of incident light of left- and right-circularpolarization states in a spectral regime called the circular Bragg regime. One wayto fabricate a chiral STF is to direct a collimated vapor flux obliquely towards auniformly rotating substrate mounted inside a vacuum chamber, the fixed anglebetween the vapor flux and the substrate plane being denoted by χv. This processis often implemented by rotating the substrate by an angle δ π and a subde-posit of height h is deposited before the next rotation by δ, resulting in a finelyambichiral STF. Another way is the asymmetric serial-bideposition (ASBD) tech-nique, whereby the substrate is rotated alternately by angles π and π + δ, δ π,and subdeposits of heights h1 and h2 = hx:1 − h1 are sequentially deposited withhx:1 = h1 + h2 fixed.

I investigated the effect on the CBP by altering the ratio h1:h2 while keeping hx:1fixed. Structurally right-handed chiral STFs of zinc selenide were deposited withratios 1:1, 1.5:1, 2:1, 2.5:1, 3:1, 5:1, 7:1, and 9:1, while hx:1 = 2.17 nm, χv = 20,and δ = 3 were kept fixed. A finely ambichiral STF was also deposited withh = 2.49 nm, χv = 20, and δ = 3. The period of all samples was fixed at 300 nmand the number of periods at 10. Measurements of the circular reflectances (RLL,RLR, RRL, and RRR) were made with the angle of incidence θinc varying between10 to 70, while measurements of the circular transmittances (TLL, TLR, TRL, andTRR) were made for θinc ∈ [0, 70]. Red-shifting and narrowing of the circularBragg regime were found to intensify with increasing h1:h2 for all values of θinc.A limit in red-shifting seemed to be achieved with the 9:1 sample, with a betterdefined circular Bragg regime than that of the finely ambichiral sample.

iii

Table of Contents

List of Figures vii

List of Tables x

Acknowledgments xii

Chapter 1Introduction 11.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.2 Timeline of STFs . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

1.2.1 Physical Vapor Deposition . . . . . . . . . . . . . . . . . . . 41.2.2 Salient Developments in CTF History . . . . . . . . . . . . . 6

1.2.2.1 Structure Zone Model . . . . . . . . . . . . . . . . 71.2.2.2 CTF Growth . . . . . . . . . . . . . . . . . . . . . 9

1.2.3 STF History and Growth . . . . . . . . . . . . . . . . . . . . 101.2.3.1 Evolution of STF Growth Methods . . . . . . . . . 11

1.3 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

Chapter 2The Circular Bragg

Phenomenon 162.1 The Bragg Phenomenon . . . . . . . . . . . . . . . . . . . . . . . . 162.2 Reusch Piles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182.3 Cholesteric Liquid Crystals . . . . . . . . . . . . . . . . . . . . . . . 192.4 Chiral STFs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

Chapter 3Experimental Procedure 223.1 PVD Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223.2 Material Choice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

iv

3.3 Substrate Preparation . . . . . . . . . . . . . . . . . . . . . . . . . 243.4 Deposition Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . 253.5 Motor Program . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273.6 Characterization . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

3.6.1 Optical Characterization . . . . . . . . . . . . . . . . . . . . 313.6.2 Cross-section Morphology . . . . . . . . . . . . . . . . . . . 34

Chapter 4Results 354.1 Optical Responses . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36

4.1.1 Density Plots of Circular Remittances . . . . . . . . . . . . 364.1.2 Flatness Measurements . . . . . . . . . . . . . . . . . . . . . 494.1.3 Full-Width Half Maximum . . . . . . . . . . . . . . . . . . . 56

4.2 Criterion-Based Performance Parameters . . . . . . . . . . . . . . . 594.3 Cross-Section Morphology . . . . . . . . . . . . . . . . . . . . . . . 62

Chapter 5Conclusion 705.1 Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72

5.1.1 Narrow-band Circular Polarization Filter . . . . . . . . . . . 725.1.2 Vertical-Cavity-Surface-Emitting Laser . . . . . . . . . . . . 725.1.3 Surface Multi-plasmonic Resonance Imaging . . . . . . . . . 735.1.4 Theoretical Model . . . . . . . . . . . . . . . . . . . . . . . 74

Appendix ARaw Data 76

Appendix BMathematica Code for

2D-Density Plots 79

Appendix CMathematica Code for

Flatness Determination 86

Appendix DMathematica Code for FWHM Plot 91

Appendix EStandard Operating Procedure 93

v

Appendix FNon-technical Abstract 95

Bibliography 98

vi

List of Figures

1.1 Different cosine distributions can be achieved using different sourcereceptacles and are also dependent of the specific type of PVD [2]. . 5

1.2 The five possible zones of PVD in the SZM. Zone M is the correctzone for growing CTFs and STFs [2]. . . . . . . . . . . . . . . . . . 9

1.3 The growing columns of a CTF casting a shadow on the left partof the substrate [2]. . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

1.4 Cross-sectional SEM images of SNTF morphologies: (a) Zig-zag;(b) C-shaped; (c) S-shaped nanocolumns [2]. . . . . . . . . . . . . . 12

1.5 Cross-sectional SEM image of a chiral STF with a very pronouncedvoid network [29]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

1.6 Four schemes of substrate rotation to grow chiral STFs. In theseschemes, the subdeposit structure is divided differently [11]. . . . . 14

3.1 Photograph of the interior of the vacuum chamber used to depositall films for this thesis. The distance from the source material tothe substrate is approximately 15 cm. . . . . . . . . . . . . . . . . . 23

3.2 The MMI Prompt commands were used in the Applied MotionSiNet Hub Programmer to pause at critical points until manualinput was given to continue. Also in the bottom left corner is theanalysis window, from which the wait times were inputed into aMathematicaTM provided in Fig. 3.3. . . . . . . . . . . . . . . . . 29

3.3 MathematicaTM program for the 3:1 film. The values of the wait-time ratio n were changed for different h1/h2 ratios. The outputsshow what values to enter for the subdeposit wait times from Figure3.2 and how many program repeats the motor needs. . . . . . . . . 30

3.4 Photograph of the setup to measure the circular transmittances ofa sample. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32

3.5 Photograph of the setup to measure the circular reflectances of asample. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32

vii

4.1 Measured circular reflectances of the 1:1 sample. . . . . . . . . . . . 374.2 Measured circular transmittances of the 1:1 sample. . . . . . . . . . 374.3 Measured circular reflectances of the 1.5:1 sample. . . . . . . . . . . 384.4 Measured circular transmittances of the 1.5:1 sample. . . . . . . . . 384.5 Measured circular reflectances of the 2:1 sample. . . . . . . . . . . . 394.6 Measured circular transmittances of the 2:1 sample. . . . . . . . . . 394.7 Measured circular reflectances of the 2.5:1 sample. . . . . . . . . . . 404.8 Measured circular transmittances of the 2.5:1 sample. . . . . . . . . 404.9 Measured circular reflectances of the 3:1 sample. . . . . . . . . . . . 414.10 Measured circular transmittances of the 3:1 sample. . . . . . . . . . 414.11 Measured circular reflectances of the 5:1 sample. . . . . . . . . . . . 424.12 Measured circular transmittances of the 5:1 sample. . . . . . . . . . 424.13 Measured circular reflectances of the 7:1 sample. . . . . . . . . . . . 434.14 Measured circular transmittances of the 7:1 sample. . . . . . . . . . 434.15 Measured circular reflectances of the 9:1 sample. . . . . . . . . . . . 444.16 Measured circular transmittances of the 9:1 sample. . . . . . . . . . 444.17 Measured circular reflectances of the finely ambichiral sample. . . . 454.18 Measured circular transmittances of the finely ambichiral sample. . 454.19 Blue-shifting of λo for the 5:1 sample at various θinc. All samples

exhibited a similar blue-shifting effect. . . . . . . . . . . . . . . . . 484.20 Plot of TRR of the 1:1 sample at θinc = 10. The wavelength and

transmittance at point 3 were found by averaging the respectivevalues at points 1 and 2. Point 5 lies at the same wavelength as point3, and the reflectance value was found by averaging the reflectancevalues of points 3 and 4. . . . . . . . . . . . . . . . . . . . . . . . . 51

4.21 H(θinc) versus θinc for all samples. . . . . . . . . . . . . . . . . . . . 554.22 gavg(θinc) versus θinc for all samples. . . . . . . . . . . . . . . . . . . 554.23 F (θinc) versus θinc for all samples. . . . . . . . . . . . . . . . . . . . 584.24 G(θinc) versus θinc for all samples. . . . . . . . . . . . . . . . . . . . 584.25 Cross-sectional FESEM image of the 1:1 sample. The on-screen

measurement tool reads 3.26 µm as the thickness. . . . . . . . . . . 634.26 Cross-sectional FESEM image of the 1.5:1 sample. The on-screen

measurement tool reads 3.27 µm as the thickness. . . . . . . . . . . 634.27 Cross-sectional FESEM image of the 2:1 sample. The on-screen

measurement tool read 3.1 µm as the thickness. . . . . . . . . . . . 644.28 Cross-sectional FESEM image of the 2.5:1 sample. The on-screen

measurement tool reads 3.26 µm as the thickness. . . . . . . . . . . 644.29 Cross-sectional FESEM image of the 3:1 sample. The on-screen

measurement tool reads 3.34 µm as the thickness. . . . . . . . . . . 65

viii

4.30 Cross-sectional FESEM image of the 5:1 sample. The on-screenmeasurement tool reads 3.34 µm as the thickness. . . . . . . . . . . 65



4.33 Cross-sectional FESEM image of the finely ambichiral sample. Theon-screen measurement tool reads 3.25 µm as the thickness. . . . . 67

4.34 Cross-sectional FESEM image of the CTF sample. Nanocolumnsgrew at approximately χ = 37. . . . . . . . . . . . . . . . . . . . . 67

ix

List of Tables

4.1 First and second subdeposit heights and their ratio. The depositionprocedure followed for samples 1:1 to 9:1 is shown in the fourth rowof Fig. 1.6. The finely ambichiral sample followed the depositionprocedure in the first row of Fig. 1.6. . . . . . . . . . . . . . . . . 35

4.2 Parameters to evaluate the quality of the CBP of the 1:1 sample asfunctions of θinc. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52

4.3 Parameters to evaluate the quality of the CBP of the 1.5:1 sampleas functions of θinc. . . . . . . . . . . . . . . . . . . . . . . . . . . . 52


4.5 Parameters to evaluate the quality of the CBP of the 2.5:1 sampleas functions of θinc. . . . . . . . . . . . . . . . . . . . . . . . . . . . 53






4.11 Criterion-based parameters of the circular Bragg regime width ofthe 1:1 sample. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60

4.12 Criterion-based parameters of the circular Bragg regime width ofthe 1.5:1 sample. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60


x

4.14 Criterion-based parameters of the circular Bragg regime width ofthe 2.5:1 sample. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60





4.19 Criterion-based parameters of the circular Bragg regime width ofthe finely ambichiral sample. . . . . . . . . . . . . . . . . . . . . . . 62

xi

Acknowledgments

First, I thank my co-advisors Dr. Akhlesh Lakhtakia and Dr. Jian Xu for teachingme the ways of a successful researcher and for their patience. I also record mythanks to Stephen E. Swiontek and Sema Erten for helping me get acquaintedwith various lab equipment and procedures. Lastly, I thank my parents for givingme many opportunities in life and for their continuous emphasis on education.

xii

Chapter 1 |

Introduction

1.1 Overview

The morphology of a sculptured thin film (STF) is engineered by manipulating

the growth direction of parallel nanocolumns during oblique-angle physical vapor

deposition (PVD). Chiral STFs are assemblies of parallel helical nanocolumns sep-

arated by a void network, and have been experimentally and theoretically proven

to exhibit the Circular Bragg Phenomenon (CBP), whereby a chiral STF of a cer-

tain structural handedness will highly reflect circular-polarized light of the same

handedness, but not circular-polarized light of the opposite handedness [1], [2].

The CBP is displayed in a spectral regime called the circular Bragg regime. Since

the conceptualization of chiral STFs, most experimental research has been focused

on light incident normally on a chiral STF so that the propagation vector of inci-

dent light is parallel to the common axis of the helical nanocolumns. If the angle

1

of incidence θinc between the propagation vector of the incident light and the com-

mon axis of the helical nanocolumns is increased, the central wavelength λBro of

the circular Bragg regime blue shifts. This has been predicted theoretically [3],

and demonstrated experimentally for zinc selenide in the visible regime [4].

Besides optical applications, chiral STFs have biomedical, electronic, and me-

chanical applications. Theoretical studies of chiral STFs confined between two

metallic planes were undertaken by Ertekin and Lakhtakia [5]. Functioning as an

optical interconnect, this structure could help increase optoelectronic circuit board

chip-density [5]. The advancement of optical interconnects along with multiplexing

is guaranteed to send computers into the exa-scale (a billion billion computations

per second) age [6].

Seto et al. were able to show that helical nanocolumns of a chiral STF (with

a capping layer to distribute the force) respond elastically to nano-indentations,

and were able to extract a spring constant and resonance frequency of the chiral

STF [7].

The inter-nanocolumnar void network produced during oblique-angle PVD pro-

vides space for infiltration with chemical and biological substances for sensing

applications by altering the effective constitutive properties of the chiral STF.

Lakhtakia conceptualized a bilayer consisting of two chiral STFs of opposite hand-

edness to sense toxic gases [8]. A clear shift in the bandwidth ∆λBro and the central

wavelength λBro of the circular Bragg regime was calculated when going from 0 to

2

50 mole/m3 of adsorbed gas [8]. Since then, sensing of various fluids has been

experimentally verified [9–12].

This chapter goes through the history of how STFs in general arose from more

simple variations of columnar thin films (CTFs), how such morphologies can be

engineered via PVD, and the motivation for the original research reported in this

thesis. The CBP exhibited by chiral STFs is discussed in Chapter 2. Chapter

3 presents the experimental methods used for growing all the samples and for

their optical and morphological characterization. The optical responses along with

measures of the quality of the circular Bragg regimes are presented in Chapter 4.

Conclusions from the obtained results are provided in Chapter 5 as well as ideas

for future work.

1.2 Timeline of STFs

Long before the advent of STFs, it was realized that the simpler CTFs could be

produced by controlling certain parameters during film growth as well as by holding

the substrate fixed at an angle with respect to the direction of a collimated vapor

flux of the material to be deposited [13]. However, before any such films could be

realized, the methods of fabricating them first needed to exist, and hence these

methods are now discussed.

3

1.2.1 Physical Vapor Deposition

PVD methods (which include sputtering, electron-beam evaporation, cathodic-arc

deposition, resistive-heating evaporation, and pulsed-laser deposition) involve the

generation of a vapor flux by raising a bulk material to a sufficiently high energy

level so that particles known as adatoms (clusters of atoms) are ejected from the

bulk material [13,14]. The goal is for the adatoms to form a collimated vapor flux

that reaches the substrate and condenses thereupon.

In order to control film morphology, adatoms need to arrive at the substrate

surface in a predictable manner. This is achieved through increasing the mean

free path of the adatoms to orders of magnitude more than the dimensions of the

PVD chamber by reaching vacuum levels of 10−3 to 10−6 Torr for resistive-heating

evaporation and 10−1 to 10−3 Torr for sputtering [2]. In addition, a high vacuum

means that fewer impurities, such as atmospheric gases, water vapor, and dust

accumulate on the surface or in the bulk of the growing film. During PVD, the

adatoms leave the receptacle containing the source (bulk) material in a cosine-

distributed fashion, as shown in Fig. 1.1. The combination of high vacuum and

cosine distribution ensures that a large fraction of the adatoms will travel in a

straight line from the source receptacle towards the substrate. This line-of-sight

characteristic is advantageous over other thin-film methods such as chemical vapor

deposition (CVD) [14]. In the case of sputtering, line-of-sight deposition is less

easily achievable since some ionized gas needs to be introduced to knock off adatoms

4

from the source material. However, using electric and magnetic fields, magnetron

sputtering can allow for ionization at lower pressures, thereby increasing the mean

free path of adatoms [2].

Since the vapor flux is not planar, there will be a slight variation in thickness

radially from the center of the substrate. In order to reduce this effect, the sub-

strate must be placed sufficiently far from the source receptacle in order to reduce

the solid angle subtended by the substrate on the receptacle, which acts as a point

source [2]. However, this must also be balanced with the fact that moving the sub-

strate farther will reduce the maximum deposition rate that can be achieved. One

benefit that oblique-angle deposited films have is that the solid angle decreases as

the angle between the average direction of the vapor flux and the substrate normal

increases, further planarizing the vapor flux.

Figure 1.1. Different cosine distributions can be achieved using different source recep-tacles and are also dependent of the specific type of PVD [2].

5

1.2.2 Salient Developments in CTF History

The optical properties of thin films were first noticed and studied by Faraday in

1857, who used thermal evaporation to make his films [15]. The first person to

seemingly notice the non-uniformity of PVD when he deposited metal on a glass

slide was Wright, who placed a metal loop about 3–4 mm in diameter a distance

of 3 mm away from a 30-mm-diameter glass slide [16]. The proximity of the loop

and its smaller size led to a large sweeping of the polar angle in order to cover the

whole substrate surface. Due to the cosine-distributed emission of adatoms with

respect to the azimuthal angle, it is no wonder that his films possessed a radial

non-uniformity. As he desired to fabricate a metal mirror, uniformity was required

and so he devised a way to move the substrate with respect to the loop in order

to average the vapor flux direction over the whole surface.

Later, in 1886, Kundt, using a similar deposition setup, confirmed the ra-

dial non-uniformity. He went one step further and conjectured that the optical

anisotropy was due to anisotropy in the way the material had been deposited [2,17].

At this time, it was hard to prove the validity of this statement, as the necessary

technologies for scanning electron microscopy (SEM) would not be available until

many decades later. Thanks to SEM, the crucial discovery of self-shadowing was

made by König and Helwig [18], [2]. They realized that under certain conditions

when the adatom mobility on the substrate surface was low, it was possible for

some features of the growing film to block other parts of the growing films (or

6

the substrate during very early growth) from the incoming vapor flux. This self-

shadowing phenomenon is possible due to the line-of-sight characteristic of PVD.

Once adatoms arrive on the surface of the substrate, the process of their stick-

ing to the film is determined by the ratio of the substrate temperature T to the

melting temperature Tm of the bulk material [2]. If the reduced temperature T/Tm

is low, the adatoms stay locally where they first arrive. At a very high reduced

temperature, the adatoms may adsorb and then desorb from the substrate sur-

face. In between these two extremes, the adatoms can adbsorb to the substrate

and then diffuse along the surface or into the bulk of the film. This ability of the

adatoms to move once adsorbed is characterized by the adatom mobility. Depend-

ing on the parameters of the deposition, these behaviors prevent the growth of a

smooth planar film and instead cause the adatoms to cluster so that the film grows

with a conical or columnar morphology. This phenomenon was first documented

by Movchan and Demchishin in 1962 [19] and its physical characterization later

improved by Thornton [20] and Messier and Troiler-McKinstry [21].

1.2.2.1 Structure Zone Model

Movchan and Demchishin created what is now known as the Structure Zone Model

(SZM), to pictorally describe the film morphology as a function of T/Tm [2], [21].

After studying thin films grown by electron-beam evaporation, Movchan and Dem-

chishin identified Zones 1, 2, and 3. At low reduced temperature (T/Tm < 0.3) [20],

7

the adatoms arrive in 1−3-nm clusters and essentially do not move after adsorp-

tion. Since there is a distribution of sizes, larger adsorbed clusters will have a

better chance of capturing incoming adatoms. This process repeats and engenders

competitive growth, whereby large clusters surpass small ones and continue to grow

in a conical form. This type of growth occurs in Zone 1 of the SZM, as illustrated

in Fig. 1.2. Zone 2 occurs at higher reduced temperatures (0.3 < T/Tm < 0.5) [20]

when the adatoms have significant surface diffusion. Here, straight-sided crystalline

grains are formed. As the temperature increases further (0.5 < T/Tm < 1) [20],

bulk diffusion into the film dominates and the columns disappear, being replaced

with equiaxed crystals. This morphology is characteristic of Zone 3.

After studying the morphology of thin films made by magnetron sputtering,

Thornton added another parameter to the SZM. This parameter was the back-

ground gas pressure, and its incorporation led to the identification of Zone T ("T"

for transistion) between Zones 1 and 2. It is here that the growth of clusters starts

to become non-competitive, and the morphology is of poorly defined fibrous grains

without void boundaries [22].

Finally, Messier and Trolier-McKinstry realized that with low ion-bombard-

ment at higher pressures, a part of Zone T gains new characteristics. Therefore

they inserted Zone M between Zones 1 and T [21]. In Zone M, the grains are

better defined, with no tapering as there was in Zone 1, but instead resemble

matchsticks. As growth occurs, ion-bombardment helps sputter adatoms to what

8

would be smaller clusters that would otherwise be shadowed by the larger clusters.

This allows for non-competitive growth of columns with a surrounding void space.

This is the desired zone for growing CTFs, and is also the appropriate zone to grow

STFs [2].

Figure 1.2. The five possible zones of PVD in the SZM. Zone M is the correct zone forgrowing CTFs and STFs [2].

1.2.2.2 CTF Growth

Prior to growing STFs, one must be able to grow CTFs. As stated in Sec. 1.2.2.1,

it was found that non-competitive columns will grow in Zone M with low-energy-

assisted ion-bombardment. The propensity toward column morphology is accentu-

ated by directing the vapor flux obliquely towards the substrate; the angle between

the vapor-flux direction and the substrate plane being denoted by χv ∈ (0, 90] in

9

Fig. 1.3 [2].

Figure 1.3. The growing columns of a CTF casting a shadow on the left part of thesubstrate [2].

Then, as clusters nucleate and grow, they block the area directly them behind

from arriving adatoms. This self-shadowing tends to work well for χv 90. The

columns grow at an angle χ often related to χv by

tan(χ) = m tan(χv), (1.1)

where m is a parameter specific to the material being deposited found from empir-

ical data [23], and χ ≥ χv [2]. The angle χ can be measured using cross-sectional

SEM images.

1.2.3 STF History and Growth

To change from columnar growth to more complicated but still locally columnar

growth, all that was needed was to instantaneously change the direction of the

vapor flux during deposition. In 1966, Nieuwenhuizen and Haanstra attempted

to grow a cascade of two CTFs with a sudden change in χv, the resulting mor-

10

phology comprising of chevronic nanocolumns [24]. This was a key step towards

the eventual emergence of STFs. Motohiro and Taga made a Ta2O5 quarter-wave

retardation film with nanocolumns of chevronic shape [25]. The difference from

Nieuwenhuizen and Haanstra’s experiment is that Motohiro and Taga changed the

vapor-flux direction several times.

However, the first to make chiral STFs were Young and Kowal [26] who de-

posited fluorite films while slowly and continuously rotating the substrate about a

central normal axis. At the time, SEM imaging was not widely available; yet they

speculated that remarkable optical properties must arise from, “helically sym-

metrical arrangement of crystallites, crystal growth or voids” [26]. From these

experiments, it was imagined that a wide range of morphologies could be possible

by changing the angle χv [2, 27]. These morphologies can fall into two categories:

nematic and helical [2]. Sculptured nematic thin films (SNTFs) consist of quasi-

2D structures such as slanted columns, chevrons, zig-zags, and C- and S-shaped

nanocolumns, as exemplified in Fig. 1.4. Thin-film helicoidal bianisotropic medi-

ums (TFHBMs) consist of nanohelixes and super-nanohelixes and are commonly

referred to as chiral STFs; see Fig. 1.5.

1.2.3.1 Evolution of STF Growth Methods

Several methods have been devised to grow complex STFs. The first attempt

was by Robbie et al. [28] who used two evaporation sources to deposit chevronic

11

Figure 1.4. Cross-sectional SEM images of SNTF morphologies: (a) Zig-zag; (b) C-shaped; (c) S-shaped nanocolumns [2].

Figure 1.5. Cross-sectional SEM image of a chiral STF with a very pronounced voidnetwork [29].

columns of SNTFs. In order to achieve vapor-flux angles χv ∈ [5, 15] the substrate

had to be placed close to the receptacle, which imposed two problems. The close

proximity to the source heated the substrate, thereby increasing adatom mobility

as became evident by the lack of porosity. The close proximity also allowed for

a greater vapor-flux gradient across the substrate, yielding a non-uniform film.

Because of these two problems, a different method was attempted. This involved

one source and rotating the substrate periodically by 180 about its central normal

axis. A large vapor source was used to ensure symmetric vapor fluxes at both

12

deposition positions. Although an improvement, these two methods were only

capable of creating SNTFs, not chiral STFs.

Unknowingly, the method of Young and Kowal [26] was rediscovered by Robbie

et al. [30] to deposit chiral STFs. Their idea was to deposit with one source and

have the substrate rotate continuously and slowly about its central normal axis.

With the ability to characterize the morphology using an SEM, they confirmed

the formation of nanohelixes, thereby vindicating Kowal and Young [26]. Instead

of continuous rotation, the substrate is often rotated in small angular increments

δ about the central normal axis, as shown in the first row of Figure 1.6. The

resulting chiral STF may be called a finely ambichiral thin film [31].

Hodgkinson et al. [32] proposed the serial bideposition (SBD) method to make

chiral STFs [32, 33], where deposits are made 180o + δ apart and the process is

repeated until a full revolution is complete, as shown in the second row of Figure

1.6. With this, Hodgkinson et al. were able to observe a significant enhancement

of the CBP at θinc = 0 exhibited by thin chiral STFs.

A further modification to the SBD method came from Pursel and Horn [11].

Two subdeposits would be made 180o from each other before the small angle δ

was introduced, as shown in the third row of Figure 1.6. With the modified SBD

method [11], the subdeposit heights are symmetric. This naturally leads to the

question of what would happen if the subdeposit heights were asymmetric, also

first explored by Pursel and Horn [11]. They compared two chiral STFs grown at

13

Figure 1.6. Four schemes of substrate rotation to grow chiral STFs. In these schemes,the subdeposit structure is divided differently [11].

χv = 12.5o and δ = 1o, the first with a 1:1 subdeposit-height ratio and the second

with a 3:1 subdeposit-height ratio. They found that their ASBD method still led to

the exhibition of the CBP by the chiral STF while possesing different void-network

characteristics. In regards to the optics of ASBD films, three main results were

found: (1) the non-co-handed transmittance of the 3:1 film was slightly less than

that of the 1:1 film, (2) the non-co-handed transmittances at lower wavelengths

of the 3:1 film was less than that of the 1:1 film, and (3) the central wavelength

red-shifted from λBro = 539 nm for the 1:1 film to λBro = 594 nm for the 3:1 film.

14

1.3 Motivation

Pursel and Horn proved that ASBD could preserve the CBP. The work done for

this thesis systematically explores the dependence of the CBP on the asymmetry

in the ASBD technique, with hopes of improving the exhibition of the CBP by

chiral STFs. Hence, chiral STFs with subdeposit-height ratios 1:1, 1.5:1, 2:1, 2.5:1,

3:1, 5:1, 7:1, and 9:1 were deposited along with a finely ambichiral sample. Fur-

thermore, this work expands upon Pursel and Horn’s research by characterizing

how the asymmetry will change the CBP for θinc > 0. Ideally, one would like

to increase the effectiveness of the CBP for circular-polarization filters for a large

range of θinc.

15

Chapter 2 |

The Circular Bragg

Phenomenon

2.1 The Bragg Phenomenon

Preceding the CBP was the Bragg phenomenon (BP), which was proposed by

William Henry Bragg and William Lawrence Bragg in 1915 [34]. The BP occurs

when electromagnetic waves within a certain spectral regime centered around a

central wavelength (λBro ) get almost totally reflected from the surface of a crystal.

The same BP is exhibited by a multi-layered material with layers of alternating

high/low indices of refraction. Each high/low-index layer needs to be one-quarter

of the wavelength desired to be highly reflected. Different bands of reflection exist

16

with their center wavelengths given by [1].

λBrom= 4Ω

mnavg cos θinc, m ∈ 1, 2, 3, ..., (2.1)

where 2Ω is the period of the index of refraction, navg is the spatially averaged

index of refraction, and θinc is the angle of incidence of the electromagnetic plane

wave with respect to the thickness direction. It is assumed for Eqn. (2.1) that the

materials are non-dispersive. For multi-layers made of isotropic materials, the BP

is insensitive to the polarization state of the incident electromagnetic wave in the

neighborhood of λBrom[1].

Discrimination between the two linearly polarized states can be forced by sim-

ply making one or both layers out of a uniaxial dielectric material [2]. The peak

reflectances for incident p- and s-polarized light are then centered around different

wavelengths. To discriminate even further, the ordinary and extraordinary refrac-

tive indices, no and ne, of the uniaxial material can be chosen such that incident

light of one state of polarization is almost totally transmitted while that of the other

is highly reflected [13]. This occurs under the conditions noa 6= nob, nea = neb

or noa = nob, nea 6= neb, where noa, nob, nea, and neb are the ordinary refractive

index of layer a, ordinary refractive index of layer b, extraordinary refractive index

of layer a, and extraordinary refractive index of layer b, respectively. One or both

layers could be made of a biaxial dielectric material. A periodic multi-layer is also

called a Bragg mirror or a Bragg reflector.

17

2.2 Reusch Piles

In a Reusch pile, a period consists of multiple layers of the same uniaxial dielectric

material such that the optic axis of every uniaxial layer is rotated in relation

to that of the previous uniaxial layer by a constant angle about the thickness

axis. It was found that a Reusch pile selectively discriminates between circularly

polarized states of normally incident light in multiple spectral regimes [1, 35–39].

For normally incident light and for the optic axis in each layer being normal to the

thickness direction, and with the assumption of the material being non-dispersive,

the circular Bragg regimes will be centered around

λBrop=

2Ω(no+ne

2 )pq + 1 , p ∈ 0, 1, 2, 3, ..., (2.2)

where q is the number of layers in a half-period, 2Ω is the period, and no and ne

are the ordinary and extraordinary refractive indices, respectively. In addition, the

uniaxial material can be replaced by a biaxial material. CTFs can be incorporated

in Reusch piles, where they have been found experimentally to exhibit two identi-

fiable circular Bragg regimes [31, 37]. As q → ∞, the changing orientation of the

permittivity dyadic from layer to layer is essentially continuous on a macroscopic

scale and the CTF can be considered to be a chiral STF [31].

18

2.3 Cholesteric Liquid Crystals

The permittivity dyadic of a chiral liquid crystal (CLC) varies continuously about

a fixed axis. CLCs possess short-range order like a liquid material, yet still possess

long-range periodicity of solid crystals [40]. CLCs fall into two categories: nematic

and smectic. Both chiral nematic liquid crystals (also called cholesteric liquid

crystals) and chiral smectic liquid crystals exhibit the CBP. These materials can

also be thought of as Reusch piles in the limit q →∞ [1].

2.4 Chiral STFs

Simlilar to CLCs, chiral STFs can be thought of as Reusch piles in the limit of

q →∞. The relative permittivity dyadic of a chiral STF is given by [2]

εr

= S· εref·S−1, (2.3)

where

εref

=

εb 0 0

0 εc 0

0 0 εa

, (2.4)

19

with εa, εb, and εc being the principal relative permittivity scalars of the biaxial

dielectric material. The helical morphology is captured by the rotation dyadic

S =

cos(hz πΩ) − sin(hz πΩ) 0

sin(hz πΩ) cos(hz πΩ) 0

0 0 1

cos(χ) 0 − sin(χ)

0 1 0

sin(χ) 0 cos(χ)

. (2.5)

Here, h can take on the value 1 or −1 for structural right or left handedness

respectively, and χ is the angle of nanocolumn growth with respect to the substrate.

Not only do εa, εb, and εc depend on the material evaporated to fabricate the chiral

STF, but also on χ and 2Ω.

The CBP for a chiral STF occurs in spectral regimes with center wavelengths

[41], [42]

λBrom≈ 2Ω

m(√εc +√εd) cos1/2 θinc, m ∈ 1, 2, 3, ..., (2.6)

where [43]

εd = εaεbεa cos2 χ+ εb sin2 χ

(2.7)

and the absence of dispersion is assumed. Only orders m > 1 will clearly distin-

guish between incident RCP and LCP plane waves. The order m = 2 is the most

important one, for which the bandwidth is

∆λBro ≈ 2Ω | √εc −√εd | cos1/2 θinc (2.8)

20

and the center wavelength

λBro ≈ Ω(√εc +√εd) cos1/2 θinc (2.9)

It is important to note that this model assumes the chiral STF to be a material

continuum; therefore it does not explicitly consider scattering of light due to the

void network.

21

Chapter 3 |

Experimental Procedure

3.1 PVD Method

The deposition method used to grow all films for this thesis was resistive-heating

thermal evaporation, which is typically used for materials with vaporizing temper-

atures below 1500 C [44]. In this method, a current is passed through a tungsten

boat containing the bulk (source) material to be evaporated. In addition to a

thermal-evaporation system being readily available at Pennsylvania State Uni-

versity, advantages of thermal evaporation include the ability to evaporate bulk

material at high vacuum, high deposition rates, relatively low power input, and

directionality. Figure 3.1 is a photograph of the vacuum chamber used.

22

Figure 3.1. Photograph of the interior of the vacuum chamber used to deposit all filmsfor this thesis. The distance from the source material to the substrate is approximately15 cm.

3.2 Material Choice

99.995% pure zinc selenide (ZnSe) supplied by Alfa Aesar® (Ward Hill, MA, USA)

was chosen for evaporation. The lumps of ZnSe were crushed into a fine powder

using a mortar and pestle. A respirator was worn during the experiments to avoid

the toxic effects of ZnSe on the respiratory system [45]. The melting point of

23

ZnSe has been reported to be 1525 C at 1 atm pressure, and will be lower under

vacuum conditions [45,46]. Kurt J. Lesker, Inc. (Jefferson Hills, PA, USA) reports

the vaporization pressure of ZnSe at 10−4 Torr to be 660 C [47]. This value will

further decrease due to the order-of-magnitude lower-vacuum conditions during

deposition, making the evaporation of ZnSe well within the capabilities of resistive-

heating thermal evaporation. ZnSe was also chosen for its high index of refraction

(n = 2.67 at λo = 550 nm) which is important for increasing the local linear

birefringence of the chiral STF [32]. Various companies report high transmission

for ZnSe over a wide range of wavelengths that cover the visible spectrum [48,49].

ZnSe is a II-VI compound semiconductor with a bandgap of 2.71 eV [50].

3.3 Substrate Preparation

Each chiral STF was deposited simultaneously on a glass substrate and silicon

substrate. VWR® pre-cleaned micro-slides (48300-0025, VWR, Radnor, PA, USA)

were cut into smaller pieces. The glass substrate was used for the sample to be

optically characterized and the silicon substrate for SEM imaging of the sample.

Both substrates were first cleaned by hand with ethanol and then bathed in ethanol

while in an ultrasonic cleaner for 5 min on each side. Immediately after removal

from the bath, the substrates were dried with pressurized nitrogen gas, secured to

the substrate holder, and placed in the vacuum chamber. Kapton® tape was used

to secure the substrates to the substrate holder.

24

3.4 Deposition Procedure

All chiral STFs were fabricated approximately 10 periods thick, each period being

approximately 300 nm. The vapor-flux incidence angle was set as χv = 20 along

with δ = 3. Due to the size limitations of the tungsten boat (S22-.005W, R. D.

Mathis, Signal Hill, CA, USA), the process was completed in three stages, each

depositing 1000 nm. In between two successive stages, the chamber was cooled and

evacuated to allow the replacement of source material. In all cases, the chamber

was allowed to cool for half-an-hour before venting to prevent thermal stresses on

the film. Care was taken not to alter the orientation of the substrate holder in

between two consecutive stages.

To monitor the deposition rate, a gold-coated quartz crystal monitor (QCM)

was used. Once placed in the probe, an oscillator circuit runs at the resonance

frequency of the crystal (6 MHz) and induces shear oscillations of the QCM volume

which in turn causes a detectable piezoelectric effect that is used to monitor the

rate. Every crystal (SPC-1093-G10, Inficon, Chicago, IL, USA) was discarded once

it reached 80% of its lifetime. To ensure the QCM was accurate, the tooling factor

was adjusted and the correct z-ratio was selected.

Since the QCM cannot be at the same position as the substrate, the QCM and

substrate will receive different amounts of the vapor flux. In order to account for

this difference, the thickness of a film must be measured by some other means,

25

such as a stylus or optical profilometer. The thickness read by the QCM is then

corrected to the measured thickness using the relation

Tf = TitmtQCM

, (3.1)

where Tf , Ti, tm, and tQCM are the final tooling factor, initial tooling factor,

measured thickness, and thickness read by the QCM, respectively.

While the tooling factor accounts for the correct amount of material on the

QCM, the z-ratio is a measure of how much the depositing material will impede the

QCM resonance oscillations. Here, z is the specific acoustic impedance (kg/m2s)

of a material, and the z-ratio is the ratio of the specific acoustic impedance of the

quartz crystal to the specific acoustic impedance of the source material. ZnSe has

a z-ratio of 0.722 [51].

Once both substrates were taped to the substrate holder which was mounted

onto the motor, the door of the vacuum chamber was closed and the chamber

was pumped down. Throughout each deposition, the pressure was kept below

10−5 Torr. Leading up to the deposition, the current was slowly turned up to

heat the tungsten boat. The current was increased at approximately 0.1 A/s

until the voltage was ∼ 2 V and the electrical power supplied to the boat was

∼ 190 W. This slow ramping rate was a safeguard to make sure that material

would not be ejected from the boat from too sudden a power increase. This also

allowed the material to outgas so that the actual deposition could take place under

26

higher vacuum. In addition, this step allowed the material to sinter homogeneously,

thereby preventing large portions of the material to sinter during deposition and

causing spikes in the deposition rate. Lastly, any contaminants on the boat or in the

source material would have had sufficient time to vaporize before the deposition.

The deposition took place at 0.4 ± 0.02 nm/s. Once this rate was reached, the

shutter was manually opened, the motor program was manually started, and the

QCM thickness reading was zeroed.

3.5 Motor Program

Two motors, the first to orient the substrate holder to the chosen value of χv and

the second to rotate the substrate about a central normal axis, were controlled

by an Applied Motion© SiNet Hub 444 Programmer (1000-267, Applied Motion

Products, Watsonville, CA, USA). Before the chamber was pumped, the first motor

was run for visual confirmation that the substrate was oriented correctly towards

the source. The second motor was programmed to automatically stop after 10/3

periods. The rotating motor must stop for each subdeposit to grow. The wait

times for consecutive subdeposits were determined by a MathematicaTM program;

a sample is shown in Fig. 3.2.

A trial run was done to ensure that the second motor stopped after the correct

amount of time and in the correct position for the next subdeposit. It was found

that the motor stopped in the correct place but would have rotated for an addi-

27

tional ∼ 1 min while 10/3 periods had been deposited. This extra time could be

attributed to two reasons. First, the SiNet Hub Programmer has one caveat, which

is that only certain values for time intervals can be inputed. Since any inputed

time interval never exactly matched the calculated time interval, any slight error

would be multiplied over the course of 400 subdeposits. The second error could

come from delay time in between motor steps. The error of ∼ 1 min for each

deposition at a rate of 0.4 nm/s would amount to a total thickness of 3075 nm.

The discrepancy of 3075 − 3000 = 75 nm is within the 5% error limit that has

been deemed acceptable from previous work [4], [52], [53]. Due to manual control

of the deposition rate during the deposition, the targeted 0.4 nm/s was not always

achieved, and therefore the thickness from sample to sample varied slightly from

3075 nm. Even with a slight error in pitch, all that matters is that each sample

was fabricated in the same fashion.

28

Figure 3.2. The MMI Prompt commands were used in the Applied Motion SiNet HubProgrammer to pause at critical points until manual input was given to continue. Alsoin the bottom left corner is the analysis window, from which the wait times were inputedinto a MathematicaTM provided in Fig. 3.3.

29

Figure 3.3. MathematicaTM program for the 3:1 film. The values of the wait-timeratio n were changed for different h1/h2 ratios. The outputs show what values to enterfor the subdeposit wait times from Figure 3.2 and how many program repeats the motorneeds.

30

3.6 Characterization

3.6.1 Optical Characterization

All samples were optically characterized within 24 h of fabrication. This was to

ensure that water adsorption and consequent degradation would not have an effect

on transmittances and reflectances. The samples were kept in a desiccator from

when they were fabricated up until the moment of characterization.

Light from a halogen source (HL-2000, Ocean Optics, Dunedin, Florida, USA)

was passed through a fiber optic cable and then through a linear polarizer (GT10,

ThorLabs, Newton, NJ, USA), a Fresnel Rhomb (LMR1, ThorLabs), the sample

to be characterized, a second Fresnel Rhomb (LMR1, ThorLabs), a second linear

polarizer (GT10, ThorLabs), and then a fiber optic cable to a CCD spectrome-

ter (HRS-BD1-025, Mightex Systems, Pleasanton, CA, USA). Measurements were

taken for the transmittance TLL, TRL, TRR, and TLR for θinc ∈ [0− 70], and for

the reflectance RLL, RRL, RRR, and RLR for θinc ∈ [10−70]. The first subscript

denotes the circular polarization state of light collected by the detector, while the

second subscript denotes the circular polarization state of light impinging on the

sample.

31

Figure 3.4. Photograph of the setup to measure the circular transmittances of a sample.

Figure 3.5. Photograph of the setup to measure the circular reflectances of a sample.

32

In order for the sample to rotate about the center, stage 1 in Figs. 3.4 and 3.5

was rotated 90 and the sample was moved until half of the initial intensity was

observed by the detector. Then, stage 1 was rotated back so that the normal to

the substrate was more or less in the direction of the incoming beam. With an

index card, the reflected light from the sample was checked against the incoming

beam. Stage 1 was then further adjusted so that these two beams overlapped.

The position of stage 1 was then zeroed. These steps ensured that the substrate

rotation and the stage rotation with respect to the incoming light would be the

same. All data were taken in a dark room to avoid noise from external sources.

The data were normalized as follows:

TLLrelative= TLLmeasured

− IdarkILL − Idark

(3.2)

TLRrelative= TLRmeasured

− IdarkILL − Idark

(3.3)

TRLrelative= TRLmeasured

− IdarkIRR − Idark

(3.4)

TRRrelative= TRRmeasured

− IdarkIRR − Idark

, (3.5)

where T can be replaced with R for the reflectance data. ILL and IRR are the

intensities of LCP and RCP, respectively, without the sample present and Idark is

the intensity in the absence of the white light source and background light sources.

33

3.6.2 Cross-section Morphology

The cross-section morphology was characterized using a FEI NovaTM NanoSEM

630 (FEI, Hillsboro, Oregon, USA). FESEM works by applying an electric field to

the tip of a sharpened tungsten filament, which allows for the emission of electrons

from a smaller area with a more narrow range of energies. This increases the

resolution compared to a thermionic emission SEM. FESEM images allowed for the

comparison of the chiral morphology and void network from sample to sample. The

freeze-fracture technique was used on films deposited on Si for edge viewing. This

technique has shown to be delicate enough to preserve biological samples [54]. All

samples were sputtered with iridium using a Quorum Emitech© K575X (Quorum

Technologies Ltd., Ashford, Kent, United Kingdom) sputter coater.

34

Chapter 4 |

Results

As mentioned in Chapter 3, nine samples were made with subdeposit heights in

the ratios described in Table 4.1. For all samples except the finely ambichiral

sample, the total height hx:1 = h1 + h2 of two consecutive subdeposits was kept

constant. For the ambichiral sample, the deposition procedure followed that of the

Ratio h1 (nm) h2 (nm)1:1 1.08 1.081.5:1 1.30 0.862:1 1.44 0.722.5:1 1.55 0.623:1 1.62 0.545:1 1.81 0.367:1 1.90 0.279:1 1.95 0.21Finely ambichiral 2.49 N/A

Table 4.1. First and second subdeposit heights and their ratio. The deposition proce-dure followed for samples 1:1 to 9:1 is shown in the fourth row of Fig. 1.6. The finelyambichiral sample followed the deposition procedure in the first row of Fig. 1.6.

first row in Figure 1.6. Since the rotation sequence of the two procedures differed,

35

the subdeposit height of the ambichiral sample had to be altered to h = 2.49 nm

in order to grow the same number of periods within the same thickness.

4.1 Optical Responses

As stated in Chapter 3, all films were illuminated by LCP and RCP plane waves

and the circular remittances were measured. The raw data were exported in an

ExcelTM worksheet. Due to the large size of each ExcelTM file, the data were

compiled using a MathematicaTM program, which is provided in Appendix B. A

sample of the raw data is provided in Appendix A. Due to the limitations of the

optical setup, circular reflectances were measured for θinc ≥ 10. Good predictions

for normal reflectance can be made from data for θinc = 10 [4].

4.1.1 Density Plots of Circular Remittances

All circular remittance were plotted as functions of λo and θinc in 2D-density plots

in Figures 4.1–4.18. For a desirable circular-polarization filter, RRR − RLL and

TLL − TRR should all be close to unity, while RRL, RLR, TRL, and TLR should be

vanishingly small. By virtue of the principle of conservation of energy, no circular

remittance can exceed unity.

36

Figure 4.1. Measured circular reflectances of the 1:1 sample.

Figure 4.2. Measured circular transmittances of the 1:1 sample.

37

Figure 4.3. Measured circular reflectances of the 1.5:1 sample.

Figure 4.4. Measured circular transmittances of the 1.5:1 sample.

38



39

Figure 4.7. Measured circular reflectances of the 2.5:1 sample.

Figure 4.8. Measured circular transmittances of the 2.5:1 sample.

40



41



42



43



44

Figure 4.17. Measured circular reflectances of the finely ambichiral sample.

Figure 4.18. Measured circular transmittances of the finely ambichiral sample.

45

Two main qualitative conclusions can be deduced from Figs. 4.1–4.16. First,

λBro redshifts for all θinc as the ratio h1/h2 increases. Second, the circular Bragg

regime narrows for all θinc with increasing ratio h1/h2.

Starting with the 1:1 sample, Figs. 4.1 and 4.2 show that the circular Bragg

regime is clearly identifiable for low values of θinc, but does not manifest clearly

for θinc & 60. This can be seen from the magnitude of the difference RRR − RLL

in the circular Bragg regime. The plots of TLR and TRL in Fig. 4.2 show that

some cross-polarization occurs on transmission in the circular Bragg regime, an

unwanted effect for circular-polarization filters.

Figures 4.3 and 4.4 for the 1.5:1 sample show considerable increase in RRR−RLL

at θinc above 60. However, discrimination between incident LCP and RCP plane

waves is still not as strong as it is at lower θinc. The difference TLL−TRR intensifies

at higher angles of incidence for the most part of the spectrum: about 20% higher

than what TLL − TRR is for the 1:1 sample.

The remittances of sample 2:1 are curious in relation to all other asymmetric

samples. This sample possesses a well-defined circular Bragg regime with an im-

provement in RRR − RLL at higher θinc compared to the 1:1 and 1.5:1 samples.

However, TLL − TRR within the circular Bragg regime is lower compared to the

1:1 and 1.5:1 samples. From the pattern formed by the other asymmetric samples,

there does not seem to be any reason why the 2:1 sample should inherently possess

decreased transmittance discrimination. This may be due to an obstruction on the

46

surface of the film during optical characterization. It was determined that it was

unnecessary to fabricate this film again, as enough data on how ASBD affects the

CBP was extracted from the other eight samples.

Red-shifting and narrowing of the circular Bragg regime becomes very clear

when comparing the 3:1 sample (Figs. 4.9 and 4.10) to the 1:1 sample (Figs.

4.1 and 4.2). The 3:1 sample continues the trend of decreased cross-polarized

transmittances in the circular Bragg regime.

From Figs. 4.11, 4.13, and 4.15, it becomes evident that the difference RRR −

RLL in the circular Bragg regime decreases as the ratio h1/h2 gets significantly

high, especially for lower θinc for the 7:1 and 9:1 samples. A slight decrease in

cross-polarization transmittance, however, continues to occur.

Since Figs. 4.13–4.16 did not differ drastically in terms of the red-shifting of the

circular Bragg regime, a limit could have been reached. To confirm this conjecture,

a finely ambichiral film was grown. The conjecture that the limit of the red-shifting

of the circular Bragg regime was reached with the 9:1 sample turned out to be

correct, since the profiles of the circular Bragg regime for the finely ambichiral

sample (Fig. 4.18) and the 9:1 sample (Fig. 4.16) overlap almost perfectly. The

difference RRR − RLL for the finely ambichiral sample is significantly weaker in

the circular Bragg regime compared to samples deposited with the ASBD method.

Figures 4.11, 4.13, 4.15, and 4.17 suggest that if the ratio h1/h2 keeps increasing

beyond 9:1, the circular Bragg regime will only discriminate strongly between RRR

47

and RLL at increasingly higher θinc. This means that an optimum value of the ratio

h1:h2 exists. In addition, λBro for all samples blue-shifts as θinc increases, which is

in agreement with previous work [3, 4].

Figure 4.19. Blue-shifting of λo for the 5:1 sample at various θinc. All samples exhibiteda similar blue-shifting effect.

48

4.1.2 Flatness Measurements

Instead of relying on color in the 2D-density plots to judge the quality of the CBP

for each sample, a quantitative approach was taken. The figure of merit

f(λo, θinc) = 2RRR(λo, θinc)−RLL(λo, θinc)RRR(λo, θinc) +RLL(λo, θinc)

(4.1)

has a value close to 2 if RRR highly dominates RLL, which means that f(λo, θinc)

can be used as a measure of discrimination by the samples between incident LCP

and RCP light.

Once the boundaries of the circular Bragg regime were found for θinc ∈ [10, 70]

at 10 intervals, the quality of the CBP was determined by two parameters. The

first parameter is the flatness of the CBP, and the second parameter is the magni-

tude of the difference between the average of f(λo, θinc) within the circular Bragg

regime and the ideal maximum value of 2.

The flatness of the CBP can be used to quantify how consistently the sample

is discriminating throughout the entire circular Bragg regime for a fixed θinc. I

defined the flatness function

H(θinc) = 1λmaxo − λmino

N∑n=1

∣∣∣f(λon , θinc)− favg(θinc)∣∣∣2(λon+1 − λon) , (4.2)

where λmaxo and λmino are the upper and lower bounds of the circular Bragg regime

49

respectively, N is the number of wavelengths λon selected for the averaging proce-

dure, and

favg(θinc) = 1N

N∑n=1

f(λon , θinc). (4.3)

Figure 4.20 shows an example of a transmittance plot used to find the lower and

upper bounds (points 1 and 2, respectively) of the circular Bragg regime for the

1:1 sample at θinc = 10. Points 1 and 2 were also found from the relfectance

plot. Points 1 and 2 from both plots were then averaged to find λmino and λmaxo ,

respectively. The plots were made using the MathematicaTM code provided in

Appendix D.

50

Figure 4.20. Plot of TRR of the 1:1 sample at θinc = 10. The wavelength and trans-mittance at point 3 were found by averaging the respective values at points 1 and 2.Point 5 lies at the same wavelength as point 3, and the reflectance value was found byaveraging the reflectance values of points 3 and 4.

The second parameter is defined as

gavg(θinc) = |favg(θinc)− 2| . (4.4)

If the CBP for a sample is ideally uniform (flat) and RRR−RLL is high throughout

the entire circular Bragg regime for a given θinc, then gavg(θinc) will be close to zero.

Tables 4.2–4.10 compile the data gathered from Eqns. (4.2) and (4.4). Figures 4.21

and 4.22 present plots of the tabled data points.

51

1:1θinc H(θinc) gavg(θinc)10 0.00131649 0.027779620 0.0171904 0.056679430 0.0389785 0.098898440 0.192069 0.25672750 0.205673 0.45050660 0.317899 0.943238

Table 4.2. Parameters to evaluate the quality of the CBP of the 1:1 sample as functionsof θinc.

1.5:1θinc H(θinc) gavg(θinc)10 0.0040231 0.031121420 0.0116021 0.050057630 0.0122508 0.072410940 0.0805326 0.19820450 0.340013 0.53454960 0.380371 0.9001

Table 4.3. Parameters to evaluate the quality of the CBP of the 1.5:1 sample asfunctions of θinc.



52

2.5:1θinc H(θinc) gavg(θinc)10 0.00396625 0.035795320 0.00331096 0.034061230 0.0187745 0.081239540 0.103959 0.21061950 0.266196 0.45165560 0.260344 0.712146

Table 4.5. Parameters to evaluate the quality of the CBP of the 2.5:1 sample asfunctions of θinc.





53





Ambichiralθinc H(θinc) gavg(θinc)10 0.295102 0.13876820 0.166418 0.10121230 0.298627 0.35181940 0.472629 0.62001350 0.409609 0.90308160 0.434223 1.35272


54

Figure 4.21. H(θinc) versus θinc for all samples.

Figure 4.22. gavg(θinc) versus θinc for all samples.

55

For any given sample, with some exceptions, there is an increase in H(θinc) and

gavg(θinc) with increasing θinc. No pattern seems to be clear between samples when

comparing H(θinc) or gavg(θinc) for any given θinc. This suggests that changing

the ratio h1:h2, at least up until 9:1, does not significantly affect the strength

or uniformity of discrimination between RRR and RLL within the circular Bragg

regime.

4.1.3 Full-Width Half Maximum

The center wavelength λBro (θinc) and the full-width half maximum (FWHM)

∆λBro (θinc) of the circular Bragg regime were also found from Figure 4.20 for θinc ∈

[10, 60] every 10. The value of λBro is the wavelength of point 3 in Fig. 4.20.

From point 3, the corresponding value of RRR (or TRR) at point 4 was found.

The values of RRR (or TRR) at point 3 and 4 were averaged to find the half-

maximum of the peak (or trough), (point 5 in Fig. 4.20). The value of ∆λBro

was then measured at point 5. With the method just described, ∆λBro (θinc) was

unobtainable at θinc = 70 for some samples. Therefore, ∆λBro (θinc) was only found

up to θinc = 60.

The values of λBro and ∆λBro from RRR and TRR were averaged and the functions

F (θinc) = 1α

λBro (θinc)cos1/2 θinc

(4.5)

and

56

G(θinc) = 1β

∆λBro (θinc)cos1/2 θinc

(4.6)

were defined. Here, α = λBr (0) and β = ∆λBro (0), both taken from the TRR

spectrums. Figures 4.23 and 4.24 show the plots of F (θinc) and G(θinc) respectively.

From Figure 4.23, it can be seen that all samples have a similar variation of λBro

with respect to θinc. Figure 4.24 suggests there is no simple relationship between

the FWHM values and the subdeposit-height ratio.

57

Figure 4.23. F (θinc) versus θinc for all samples.

Figure 4.24. G(θinc) versus θinc for all samples.

58

4.2 Criterion-Based Performance Parameters

If RLL ≤ 0.05RRR for satisfactory performance as a circular-polarization filter, then

f(λo, θinc) ≥ 1.8. The enforcement of this criterion means that the circular Bragg

regime fails to discriminate strongly enough for θinc > 40 for all nine samples.

With this criterion, the circular Bragg regimes used to calculate the data in Tables

4.2–4.10 were further refined for each θinc. Tables 4.11–4.19 present λmino , λmaxo ,

and δλBro = λmaxo − λmino as functions of θinc ∈ [10, 40] for all nine samples. The

spectral width δλBro of the circular Bragg regime narrows as θinc increases for every

sample. In addition, there is a trend (with some exceptions) of smaller δλBro as

h1/h2 increases. In some instances, the criterion f(λo, θinc) ≥ 1.8 yielded a circular

Bragg regime wider than shown in Figs. 4.1–4.18. In such cases, λmino and λmaxo

were obtained directly by the process shown in Fig. 4.20.

59

1:1θinc λmino (nm) λmaxo (nm) δλBr (nm)10 587.05 652.6 65.5520 582.15 648.3 66.1530 574 631 5740 566 617 51

Table 4.11. Criterion-based parameters of the circular Bragg regime width of the 1:1sample.

1.5:1θinc λmino (nm) λmaxo (nm) δλBr (nm)10 596.2 664.2 6820 594 656 6230 585 643 5840 575 628 53

Table 4.12. Criterion-based parameters of the circular Bragg regime width of the 1.5:1sample.

2:1θinc λmino (nm) λmaxo (nm) δλBr (nm)10 584.6 649.5 64.920 580.9 642.8 61.930 576 631 5540 571 616 45


2.5:1θinc λmino (nm) λmaxo (nm) δλBr (nm)10 598 663 6520 595 656 6130 588 643 5540 582 628 46

Table 4.14. Criterion-based parameters of the circular Bragg regime width of the 2.5:1sample.

60

3:1θinc λmino (nm) λmaxo (nm) δλBr (nm)10 603 667.3 64.320 598 660 6230 592 647 5540 585 634 49


5:1θinc λmino (nm) λmaxo (nm) δλBr (nm)10 609 668 5920 602 661 5930 595 650 5540 589 633 44


7:1θinc λmino (nm) λmaxo (nm) δλBr (nm)10 609.1 667.9 58.220 603 660.55 57.5530 595 646 5140 590 629 39


9:1θinc λmino (nm) λmaxo (nm) δλBr (nm)10 615 672 5420 609 668 5930 601 651 5040 596 627 31


61

Ambichiralθinc λmino (nm) λmaxo (nm) δλBr (nm)10 620.054 654.25 34.19620 615.386 650.885 35.49930 607.343 636.901 29.55840 606.046 621.35 15.304

Table 4.19. Criterion-based parameters of the circular Bragg regime width of the finelyambichiral sample.

4.3 Cross-Section Morphology

Thickness measurements taken by the FESEM tool revealed that the samples

ranged from 3.1 to 3.34 µm in thickness for reasons mentioned in Sec. 3.5. To

get a clear image of morphology within the sample, away from any edge-growth

effects, the samples were cleaved using the freeze-fracture technique. In addition,

a CTF was grown and the angle χ measured, which was 37. This is consistent

with previous findings that columns tend to grow towards the normal with respect

to χv [2].

62

Figure 4.25. Cross-sectional FESEM image of the 1:1 sample. The on-screen measure-ment tool reads 3.26 µm as the thickness.

Figure 4.26. Cross-sectional FESEM image of the 1.5:1 sample. The on-screen mea-surement tool reads 3.27 µm as the thickness.

63

Figure 4.27. Cross-sectional FESEM image of the 2:1 sample. The on-screen measure-ment tool read 3.1 µm as the thickness.

Figure 4.28. Cross-sectional FESEM image of the 2.5:1 sample. The on-screen mea-surement tool reads 3.26 µm as the thickness.

64



65



66

Figure 4.33. Cross-sectional FESEM image of the finely ambichiral sample. The on-screen measurement tool reads 3.25 µm as the thickness.

Figure 4.34. Cross-sectional FESEM image of the CTF sample. Nanocolumns grew atapproximately χ = 37.

67

Helical nanocolumns are clearly seen in the 1:1 sample (Fig. 4.25), rising per-

pendicularly to the substrate. Even with a slight asymmetry, as in the 1.5:1 sample

(Fig. 4.26), the nanocolumns look quite different compared to the 1:1 sample. Each

nanocolumn takes on a C-shape that grows tilted locally, yet over 10 periods stays

upright. The C-shape becomes better defined with increasing ratio h1/h2. As a

thought experiment, one can imagine that for any ratio h1/h2 there will initially

be tilting since one sub-deposit is shorter than the other. After one half of a period

has been deposited, the longer sub-deposit will be on the other side making the

structure lean the other way. Hence, after one period, the structure will overall stay

perpendicular to the substrate. The 3:1 sample (Fig. 4.29), however, does seem to

lean somewhat. To verify whether this was due to cleaving or some unaccounted

factor during growth, a 5:1 sample was made. From the FESEM micrograph in

Fig. 4.30, all helical nanocolumns look relatively straight, eliminating any previous

doubts from the 3:1 sample. For the 5:1 sample, the deposition time for h2 was

calculated to be ∼ 0.905s, while the time to rotate 180 and 180+ δ for each sub-

deposit was on the order of ∼ 0.41s. Since the 5:1 sample still produced a strong

CBP, I decided that the limit should be pushed even further. Therefore, samples

with ratios of 7:1 (Fig. 4.31) and 9:1 (Fig. 4.32) were made. At this point, it was

expected that a limit would be reached. Since the shorter sub-deposit becomes

increasingly insignificant as h1/h2 is increased, at some point, the film essentially

becomes a finely ambichiral STF with one sub-deposit, as in the method described

68

in the first row of Fig. 1.6.

Therefore, an ambichiral STF was grown for comparison and is shown in Figure

4.33. Figures 4.31, 4.32, and 4.33, respectively, show that the morphologies of the

7:1, 9:1 and the finely ambichiral samples are indistinguishable. For the 7:1 and

9:1 samples, the deposition heights for the shorter subdeposit was ∼ 0.27nm and

∼ 0.21 nm, respectively. There heights are comparable to the ∼ 0.16 nm deposited

each turn. Even though h2 did not seem to have an effect on growth, it is clearly

still important for increasing the local linear birefringence as seen by the higher

discrimination (i.e. RRR−RLL) in the circular Bragg regime when comparing Figs.

4.13–4.16 to Figs. 4.17 and 4.18.

69

Chapter 5 |

Conclusion

I set out to find how ASBD chiral STFs affect the CBP with light impinging at

angle of incidence θinc ∈ [0, 70]. The films were fabricated with ZnSe as the

source material, due to its high index of refraction, optical transparency in the

visible regime, and low-temperature vaporization at pressures below 10−4 Torr.

Evaporation is easily achievable using resistive-heating PVD. Other benefits from

this fabrication method include line-of-sight growth and a high as well as stable

deposition rate.

Inside the deposition chamber, two motors were used to orient the substrate:

one to the ensure that the collimated vapor flux was directed at the chosen value

of χv (= 20) with respect to the substrate plane, and the other to rotate the

substrate about its central normal axis. For all samples, the substrate was rotated

by 180 between the first subdeposit of height h1 and the second subdeposit h2, and

by 180 + δ between the subdeposit of height h2 and the subdeposit of height h1.

70

The ratio of the two subdeposit heights h1/h2 was changed to create asymmetric

growth, with δ = 3 fixed. Chiral STFs were deposited for ratios h1:h2 of 1:1, 1.5:1,

2:1, 2.5:1, 3:1, 5:1, 7:1, and 9:1. Their optical characteristics were compared to

those of a finely ambichiral STF (for which the ratio h1/h2 →∞). First, LCP light

was incident and the LCP and RCP transmittances and reflectances were measured

as functions of angle of incidence θinc and free-space wavelength λo. Then, RCP

light was incident and the LCP and RCP transmittances and reflectances were

similarly measured. Furthermore, the morphology of each sample was studied

using FESEM images.

Red-shifting of the central wavelength λBro of the circular Bragg regime for all

θinc ∈ [0, 70] and narrowing of a criterion-based bandwidth δλBro of the circular

Bragg regime for all θinc ∈ [0, 40] with increasing h1/h2 were the two significant

findings. If RLL was set to not exceed 5% of RRR, then all samples performed

satisfactory only for θinc ≤ 40. Morphologically, the second subdeposit heights

for samples 7:1 and 9:1 were so small that both of these samples were virtually

indistinguishable from the finely ambichiral sample. The major conclusions of the

research were as follows: chiral STFs grown using the ASBD method are more

suitable than SBD chiral STFs for narrow-band circular polarization filters, ASBD

chiral STFs are more promising as vertical-cavity-surface-emitting laser mirrors,

and ASBD chiral STFs may be beneficial in surface multi-plasmonic resonance

imaging.

71

5.1 Future Work

5.1.1 Narrow-band Circular Polarization Filter

This work has shown that the δλBro can be reduced by 17% at θinc = 10 and by

39% at θinc = 40, if a 9:1 chiral STF is substituted by a 1:1 chiral STF. This would

make chiral STFs grown using the ASBD method more suitable than SBD chiral

STFs for narrow-band circular polarization filters. As mentioned in Sec. 4.1.1, the

difference RRR − RLL decreases slightly from the 7:1 sample to the 9:1 sample.

The film with the best balance of narrow bandwidth and high RRR −RLL is most

likely between these two ratios.

5.1.2 Vertical-Cavity-Surface-Emitting Laser

One successful application of asymmetric chiral STFs of high h1/h2 ratio could

be in circularly polarized lasers as cavity mirrors. The concept and realization of

chiral-mirror-based micro-cavity light-emitting diodes and vertical-cavity-surface-

emitting lasers has already been shown to work for symmetric chiral STFs [56],

[57]. The center wavelength of the laser could be tuned by appropriately changing

h1 +h2 = h and/or by post-deposition annealing [58] and/or chemical etching [59].

72

5.1.3 Surface Multi-plasmonic Resonance Imaging

Surface multi-plasmonic resonance imaging (SMPRI) is an emerging technique to

optically sense analytes in aqueous solutions [12]. A beam of light is incident

upon one slanted face of a glass prism, on whose base a metal film and a chiral

STF are successively deposited. The beam gets reflected at the base of the prism

and emerges from the other slanted face of the prism to be captured by a charge-

coupled device (CCD) spectrometer. There will be a range of angles of incidence for

which surface-plasmon-polariton (SPP) waves will be excited to propagate along

the interface of the metal and the chiral STF. When an SPP wave is excited, a

sudden decrease in the light intensity measured by the spectrometer will occur. In

order to sense analytes, in aqueous solution, the void network of the chiral STF

can be infiltrated by the solution to change the dielectric properties of the chiral

STF, and thus the angles of incidence for which a sudden decrease in the measured

intensity occurs. Depositing the chiral STF using the SBD method can result in an

increase in local linear birefringence, but a decreased void-network volume for the

aqueous solution to penetrate. Use of a finely ambichiral STF allows for a larger

void-network volume, but comes at the cost of decreased birefringence. Chiral

STFs deposited using the ASBD method could provided a balance between these

two characteristics.

73

5.1.4 Theoretical Model

Modeling can help pin-point interesting effects that may have been missed in this

research because only a small number of ASBD chiral STFs were fabricated for

experiments. Asymmetric growth is not represented by the rotation dyadic pre-

sented in Eqn. (2.5). Another model suggested by Venugopal [60] could be used

to describe the asymmetric chiral structures. His model consists of creating a lat-

tice with periodicity in two mutually orthogonal directions. At the center of each

lattice point, a thin cylinder is placed. Venugopal’s model is described by the

convolution of a cylinder function with two Dirac-delta functions, each controlling

the periodicity in the x and y directions respectively. The lattice is then made to

rotate as it rises in the z direction with the functions, thus,

xa(z) = xa(0) +R sin πzΩ , (5.1)

ya(z) = ya(0) +R(

1− cos πzΩ

), (5.2)

where 2Ω is one period, R is the helical sweep radius, and xa(0) and ya(0) are the

positions of the centers of the cylinders in the plane z = 0. These equations develop

the thin cylinders in a helical fashion as z increases. For asymmetric growth, xa(z)

and ya(z) need to be changed appropriately. The optical characteristics can be

computed using rigorous coupled-wave analysis [60]. This model also has the added

74

benefit of accounting for the 3D periodicity of chiral STFs.

75

Appendix A|

Raw Data

Reflectance data for the 1:1 sample when θinc = 10 is shown to provide an exam-

ple of how the raw optical intensity data was laid out in Microsoft Excel. Since

the data spanned 3717 rows, only the beginning and the end are shown here to

give an idea of how the meaningful data were picked out with MathematicaTM code.

a = Import[

"C:\\Penn State\\Research\\PatM_ZnSe_ASBD\\GOOD DATA\\

ZnSe 300nm

1_1 (10_26_2015)\\RLL\\10.csv"];

Flatten[Table[a[[1 + (3717− 52)i;;52 + (3717− 52)i, 1;;5]], i, 0, 1],

1]@mightex_spectrum_save, 2, 1, 0, , 10/26/2015 12:24, , , , , HRS-BD1-025,

, , , , 08-120111-007, , , , , 100, , , , , 0, , , , , 3648, , , , , 10000, , , , ,

1170., , , , ,

−0.249, , , , , −2.85*∧-6, , , , , 1.54*∧-10, , , , , 4520., , , , , −3.77, , , , ,

76

−0.0000325, , , , , −3.99*∧-8, , , , , uWcmnm,

, , , , PixelNo,Wavelength,CCD_1,CCD_2,CCD_3, 0, 1170., 0.001, 0., 0.,

1, 1170., 2.1, 0., 0., 2, 1170., 0.001, 0., 0., 3, 1170., 0.6, 0., 0.,

, 4, 1170., 0.001, 0., 0., 5, 1170., 4.6, 0., 0., 6, 1170., 0.001, 0., 0.,

7, 1170., 0.1, 0., 0., 8, 1170., 0.001, 0., 0., 9, 1170., 11.9, 0., 0.,

10, 1170., 0.3, 0., 0., 11, 1170., 0.001, 0., 0., 12, 1170., 0.001, 0., 0.,

13, 1170., 0.8, 0., 0., 14, 1170., 20.3, 0., 0., 15, 1170., 0.001, 0., 0.,

16, 1170., 0.001, 0., 0., 17, 1170., 0.001, 0., 0., 18, 1170., 4.7, 0., 0.,

19, 1160., 3.1, 0., 0., 20, 1160., 10.2, 0., 0., 21, 1160., 0.4, 0., 0.,

22, 1160., 1.8, 0., 0., 23, 1160., 12., 0., 0., 24, 1160., 0.001, 0., 0.,

25, 1160., 0.001, 0., 0., 26, 1160., 0.001, 0., 0., 27, 1160., 0.001, 0., 0.,

28, 1160., 8.6, 0., 0., 29, 1160., 9.6, 0., 0., 30, 1160., 0.001, 0., 0.,

31, 1160., 0.5, 0., 0., 32, 1160., 0.001, 0., 0., 33, 1160., 4.5, 0., 0.,

3647, 231., 11.9, 0., 0., CIE Parameters, , , , , CIEStandardObserver, 0, , , ,

DrkSub 1931x, 0, , , , drkSub 1931y, 0, , , , drkSub 1976u’, 0, , , ,

drkSub 1976v’, 0, , , , CCT of StandardLight Source,Equal Energy, , , ,

TvR 1931x, 0, , , , TvR 1931y, 0, , , , TvR 1976u’, 0, , , , TvR 1976v’, 0, , , ,

TvR 1976L*, 0, , , , TvR 1976a*, 0, , , , TvR 1976b*, 0, , , ,

Std Src 1931x, 0, , , , Std Src 1931y, 0, , , , TvRStartWav, 230.611, , , ,

TvREndWav, 1169.73, , , , AbsInten 1931x, 0, , , , AbsInten 1931y, 0, , , ,

AbsInten 1976u’, 0, , , , AbsInten 1976v’, 0, , , , CCT, 0, , , , CRIa, 0, , , ,

CRIs 1, 0, , , , CRIs 2, 0, , , , CRIs 3, 0, , , , CRIs 4, 0, , , , CRIs 5, 0, , , ,

CRIs 6, 0, , , , CRIs 7, 0, , , , CRIs 8, 0, , , , CRIs 9, 0, , , , CRIs 10, 0, , , ,

CRIs 11, 0, , , , CRIs 12, 0, , , , CRIs 13, 0, , , , CRIs 14, 0, , , ,

Irradiance, 0, , , , Illuminance, 0, , , , Purity, 0, , , , DominantWav, 0, , , ,

AbsStartWav, 230.611, , , , AbsEndWav, 1169.73, , , ,

77

MonoStartWav, 230.611, , , , MonoEndWav, 1169.73, , , , PeakWav, 0, , , ,

FWHMWav, 0, , , , CenterWav, 0, , , , CentroidWav, 0, , , ,

User Comments, , , ,

78

Appendix B|

Mathematica Code for

2D-Density Plots

The code shown in this appendix was used for making the 2D-density plot of the

3:1 sample shown in Figs. 4.9 and 4.10. Similar code was used to create the

2D-density plots in Figs. 4.1-4.18. Subtracting out dark intensities from the re-

mittance intensities resulted in some small negative values due to noise. This was

fixed by imposing an ‘IF’ statement so that intensity values below zero were given

a value of 0.

79

rmin = 1088;

rmax = 2641;

Darkdata1 =

Import["C:\\Penn State\\Research\\PatM_ZnSe_ASBD\\GOOD DATA\\ZnSe 300nm

3_1 (11_5_2015)\\Dark_Data_3_1.csv"];

Darkdata2 = Darkdata1[[rmin ;; rmax, 3]];

ALLdata1 =

Import["C:\\Penn State\\Research\\PatM_ZnSe_ASBD\\GOOD DATA\\ZnSe 300nm 3_1

(11_5_2015)\\ALL_3_1.csv"];

ALLdata = ALLdata1[[rmin ;; rmax, 3]];

ARRdata1 =


(11_5_2015)\\ARR_3_1.csv"];

ARRdata = ARRdata1[[rmin ;; rmax, 3]];

λ1 = Import["C:\\Penn State\\Research\\PatM_ZnSe_ASBD\\GOOD

DATA\\ZnSe 300nm 3_1 (11_5_2015)\\Dark_Data_3_1.csv"];

λ = λ1[[rmin ;; rmax, 2]];

θ = Flatten[0, Table[i, i, 10, 70, 2]];

Printed by Wolfram Mathematica Student Edition

80

RLLfiles = Table[


(11_5_2015)\\RLL\\" <> ToString[i] <> ".csv", "Data"], i, 10, 70, 2];

RLRfiles = Table[


(11_5_2015)\\RLR\\" <> ToString[i] <> ".csv", "Data"], i, 10, 70, 2];

RRLfiles = Table[


(11_5_2015)\\RRL\\" <> ToString[i] <> ".csv", "Data"], i, 10, 70, 2];

RRRfiles = Table[


(11_5_2015)\\RRR\\" <> ToString[i] <> ".csv", "Data"], i, 10, 70, 2];

TLLfiles =


3_1 (11_5_2015)\\TLL\\0.csv"],

Flatten[Table[Import["C:\\Penn State\\Research\\PatM_ZnSe_ASBD\\GOOD

DATA\\ZnSe 300nm 3_1 (11_5_2015)\\TLL\\" <>

ToString[i] <> ".csv", "Data"], i, 10, 70, 2], 1];

TLRfiles =


3_1 (11_5_2015)\\TLR\\0.csv"],


DATA\\ZnSe 300nm 3_1 (11_5_2015)\\TLR\\" <>


TRLfiles =


3_1 (11_5_2015)\\TRL\\0.csv"],


DATA\\ZnSe 300nm 3_1 (11_5_2015)\\TRL\\" <>


TRRfiles =


3_1 (11_5_2015)\\TRR\\0.csv"],


DATA\\ZnSe 300nm 3_1 (11_5_2015)\\TRR\\" <>


2 3_1_Density_Plots.nb


81

Rll = Flatten[RLLfiles, 1];

Rlr = Flatten[RLRfiles, 1];

Rrl = Flatten[RRLfiles, 1];

Rrr = Flatten[RRRfiles, 1];

Tll = Flatten[TLLfiles, 1];

Tlr = Flatten[TLRfiles, 1];

Trl = Flatten[TRLfiles, 1];

Trr = Flatten[TRRfiles, 1];

RLL = Transpose[Table[Rll[[rmin + 3717 i ;; rmax + 3717 i, 3]], i, 0, 30, 1]];

RLR = Transpose[Table[Rlr[[rmin + 3717 i ;; rmax + 3717 i, 3]], i, 0, 30, 1]];

RRL = Transpose[Table[Rrl[[rmin + 3717 i ;; rmax + 3717 i, 3]], i, 0, 30, 1]];

RRR = Transpose[Table[Rrr[[rmin + 3717 i ;; rmax + 3717 i, 3]], i, 0, 30, 1]];

TLL = Transpose[Table[Tll[[rmin + 3717 i ;; rmax + 3717 i, 3]], i, 0, 31, 1]];

TLR = Transpose[Table[Tlr[[rmin + 3717 i ;; rmax + 3717 i, 3]], i, 0, 31, 1]];

TRL = Transpose[Table[Trl[[rmin + 3717 i ;; rmax + 3717 i, 3]], i, 0, 31, 1]];

TRR = Transpose[Table[Trr[[rmin + 3717 i ;; rmax + 3717 i, 3]], i, 0, 31, 1]];

a = Table[RLL[[All, i]] - Darkdata2, i, 1, 31];

A = Table[If[Transpose[a][[i, n]] > 0, Transpose[a][[i, n]], 0],

i, 1, rmax - rmin + 1, 1, n, 1, 31, 1];

B = ALLdata - Darkdata2;

rll =A

B;

RLL1 = Flatten[Table[λ[[i]], θ[[n + 1]], Part[Part[rll, i], n],

i, 1, rmax - rmin + 1, 1, n, 1, 31, 1], 1];

f = Table[RLR[[All, i]] - Darkdata2, i, 1, 31];

F = Table[If[Transpose[f][[i, n]] > 0, Transpose[f][[i, n]], 0],

i, 1, rmax - rmin + 1, 1, n, 1, 31, 1];

G = ALLdata - Darkdata2;

rlr =F

G;

RLR1 = Flatten[Table[λ[[i]], θ[[n + 1]], Part[Part[rlr, i], n],

i, 1, rmax - rmin + 1, 1, n, 1, 31, 1], 1];

h = Table[RRL[[All, i]] - Darkdata2, i, 1, 31];

H = Table[If[Transpose[h][[i, n]] > 0, Transpose[h][[i, n]], 0],

i, 1, rmax - rmin + 1, 1, n, 1, 31, 1];

J = ARRdata - Darkdata2;

rrl =H

J;

RRL1 = Flatten[Table[λ[[i]], θ[[n + 1]], Part[Part[rrl, i], n],

i, 1, rmax - rmin + 1, 1, n, 1, 31, 1], 1];

3_1_Density_Plots.nb 3


82

k = Table[RRR[[All, i]] - Darkdata2, i, 1, 31];

K = Table[If[Transpose[k][[i, n]] > 0, Transpose[k][[i, n]], 0],

i, 1, rmax - rmin + 1, 1, n, 1, 31, 1];

L = ARRdata - Darkdata2;

rrr =K

L;

RRR1 = Flatten[Table[λ[[i]], θ[[n + 1]], Part[Part[rrr, i], n],

i, 1, rmax - rmin + 1, 1, n, 1, 31, 1], 1];

m = Table[TLL[[All, i]] - Darkdata2, i, 1, 32];

M = Table[If[Transpose[m][[i, n]] > 0, Transpose[m][[i, n]], 0],

i, 1, rmax - rmin + 1, 1, n, 1, 32, 1];

P = ALLdata - Darkdata2;

tll =M

P;

TLL1 = Flatten[Table[λ[[i]], θ[[n]], Part[Part[tll, i], n],

i, 1, rmax - rmin + 1, 1, n, 1, 32, 1], 1];

q = Table[TLR[[All, i]] - Darkdata2, i, 1, 32];

Q = Table[If[Transpose[q][[i, n]] > 0, Transpose[q][[i, n]], 0],

i, 1, rmax - rmin + 1, 1, n, 1, 32, 1];

R = ALLdata - Darkdata2;

tlr =Q

R;

TLR1 = Flatten[Table[λ[[i]], θ[[n]], Part[Part[tlr, i], n],

i, 1, rmax - rmin + 1, 1, n, 1, 32, 1], 1];

s = Table[TRL[[All, i]] - Darkdata2, i, 1, 32];

S = Table[If[Transpose[s][[i, n]] > 0, Transpose[s][[i, n]], 0],

i, 1, rmax - rmin + 1, 1, n, 1, 32, 1];

T = ARRdata - Darkdata2;

trl =S

T;

TRL1 = Flatten[Table[λ[[i]], θ[[n]], Part[Part[trl, i], n],

i, 1, rmax - rmin + 1, 1, n, 1, 32, 1], 1];

u = Table[TRR[[All, i]] - Darkdata2, i, 1, 32];

U = Table[If[Transpose[u][[i, n]] > 0, Transpose[u][[i, n]], 0],

i, 1, rmax - rmin + 1, 1, n, 1, 32, 1];

V = ARRdata - Darkdata2;

trr =U

V;

TRR1 = Flatten[Table[λ[[i]], θ[[n]], Part[Part[trr, i], n],

i, 1, rmax - rmin + 1, 1, n, 1, 32, 1], 1];



83

LL = ListDensityPlot[RLL1, PlotRange → All, All, All,

BaseStyle → FontWeight → Plain, FontFamily → "Helvetica",

FrameStyle → Directive[20, Plain], Frame → True, FrameTicks →

10, 20, 30, 40, 50, 60, 70, None, 500, 600, 700, 800, 900, None,

FrameLabel → Style["λo (nm)", 30, Plain],

Style["θ (deg)", 30, Plain], Style["RLL", 30, Plain],

ColorFunction → "Rainbow", ColorFunctionScaling → False];

LR = ListDensityPlot[RLR1, PlotRange → All, All, All,



10, 20, 30, 40, 50, 60, 70, None, 500, 600, 700, 800, 900, None,


Style["θ (deg)", 30, Plain], Style["RLR", 30, Plain],


RL = ListDensityPlot[RRL1, PlotRange → All, All, All,



10, 20, 30, 40, 50, 60, 70, None, 500, 600, 700, 800, 900, None,


Style["θ (deg)", 30, Plain], Style["RRL", 30, Plain],


RR = ListDensityPlot[RRR1, PlotRange → All, All, All,



10, 20, 30, 40, 50, 60, 70, None, 500, 600, 700, 800, 900, None,


Style["θ (deg)", 30, Plain], Style["RRR", 30, Plain],


Legended[GraphicsGrid[LL, LR, RL, RR, ImageSize → Scaled[.45]], BarLegend[

"Rainbow", 0, 1, LegendLayout → "Column", LabelStyle → FontSize → 20]]

3_1_Density_Plots.nb 5


84

LL1 = ListDensityPlot[TLL1, PlotRange → All, All, All,



0, 10, 20, 30, 40, 50, 60, 70, None, 500, 600, 700, 800, 900, None,


Style["θ (deg)", 30, Plain], Style["TLL", 30, Plain],


LR1 = ListDensityPlot[TLR1, PlotRange → All, All, All,



0, 10, 20, 30, 40, 50, 60, 70, None, 500, 600, 700, 800, 900, None,


Style["θ (deg)", 30, Plain], Style["TLR", 30, Plain],


RL1 = ListDensityPlot[TRL1, PlotRange → All, All, All,



0, 10, 20, 30, 40, 50, 60, 70, None, 500, 600, 700, 800, 900, None,


Style["θ (deg)", 30, Plain], Style["TRL", 30, Plain],


RR1 = ListDensityPlot[TRR1, PlotRange → All, All, All,



0, 10, 20, 30, 40, 50, 60, 70, None, 500, 600, 700, 800, 900, None,


Style["θ (deg)", 30, Plain], Style["TRR", 30, Plain],


Legended[GraphicsGrid[LL1, LR1, RL1, RR1, ImageSize → Scaled[.45]], BarLegend[

"Rainbow", 0, 1, LegendLayout → "Column", LabelStyle → FontSize → 20]]



85

Appendix C|

Mathematica Code for

Flatness Determination

This code was used to evaluate the right sides of Eqns. (4.2) and (4.4) of Sec.

4.1.2. This specific example is for the 9:1 sample. The matrix ‘u’ consists of

the wavelengths outputted by the CCD spectrometer along with their f(λo, θinc)

values. Cases where f(λo, θinc) < 1.8 were made to equal 0 so that the circular

Bragg regime could easily be picked out by visual inspection.

86

In[16]:= rmin = 19;

rmax = 3666;

Darkdata1 =


9_1 (1_6_2016)\\Dark_Data_9_1.csv"];

Darkdata2 = Darkdata1[[rmin ;; rmax, 3]];

ALLdata1 =


(1_6_2016)\\ALL_9_1.csv"];

ALLdata = ALLdata1[[rmin ;; rmax, 3]];

ARRdata1 =


(1_6_2016)\\ARR_9_1.csv"];

ARRdata = ARRdata1[[rmin ;; rmax, 3]];

λ1 = Import["C:\\Penn State\\Research\\PatM_ZnSe_ASBD\\GOOD

DATA\\ZnSe 300nm 9_1 (1_6_2016)\\Dark_Data_9_1.csv"];

λ = λ1[[rmin ;; rmax, 2]];

Δλ = Table[λ[[i]] - λ[[i + 1]], i, 1, rmax - rmin, 1];


87

In[27]:=

RLLfiles = Table[


(1_6_2016)\\RLL\\" <> ToString[i] <> ".csv", "Data"], i, 10, 70, 2];

RRRfiles = Table[Import["C:\\Penn State\\Research\\PatM_ZnSe_ASBD\\GOOD

DATA\\ZnSe 300nm 9_1 (1_6_2016)\\RRR\\" <>

ToString[i] <> ".csv", "Data"], i, 10, 70, 2];

Rll = Flatten[RLLfiles, 1];

Rrr = Flatten[RRRfiles, 1];

RLL = Transpose[Table[Rll[[rmin + 3717 i ;; rmax + 3717 i, 3]], i, 0, 30, 1]];

RRR = Transpose[Table[Rrr[[rmin + 3717 i ;; rmax + 3717 i, 3]], i, 0, 30, 1]];

a1 = Table[RLL[[All, i]] - Darkdata2, i, 1, 31, 5];

A1 = Table[If[Transpose[a1][[i, n]] <= 0, .00001, Transpose[a1][[i, n]]],

i, 1, rmax - rmin + 1, 1, n, 1, 7, 1];

B1 = ALLdata - Darkdata2;

rll =A1

B1;

a2 = Table[RRR[[All, i]] - Darkdata2, i, 1, 31, 5];

A2 = Table[If[Transpose[a2][[i, n]] <= 0, .00001, Transpose[a2][[i, n]]],

i, 1, rmax - rmin + 1, 1, n, 1, 7, 1];

B2 = ARRdata - Darkdata2;

rrr =A2

B2;

z =rrr - rll

.5 (rrr + rll);

In[64]:= z1 = Table[λ[[i]], z[[i, 1]], i, 1, rmax - rmin + 1, 1];

z2 = Table[λ[[i]], z[[i, 2]], i, 1, rmax - rmin + 1, 1];






MatrixForm[z1];

y = Table[If[z1[[i, 2]] < 1.8, 0, z1[[i, 2]]], i, 1, Length[z1]];

u = Table[λ[[i]], y[[i]], i, 1, Length[z1]];

MatrixForm[u];

2 1_1_Flatness.nb


88

rmin1 = 1963;

rmax1 = 2192;

θ10 = Transpose[λ[[rmin1 - 18 ;; rmax1 - 18]], z[[rmin1 - 18 ;; rmax1 - 18, 1]]];

rmin2 = 1988;

rmax2 = 2211;


rmin3 = 2024;

rmax3 = 2251;


rmin4 = 2081;

rmax4 = 2298;


rmin5 = 2142;

rmax5 = 2340;


rmin6 = 2199;

rmax6 = 2386;


avg10 = Sumθ10[[n, 2]]

Length[θ10], n, 1, Length[θ10];

avg20 = Sumθ20[[n, 2]]


avg30 = Sumθ30[[n, 2]]


avg40 = Sumθ40[[n, 2]]


avg50 = Sumθ50[[n, 2]]


avg60 = Sumθ60[[n, 2]]


1_1_Flatness.nb 3


89

flat10 =1

λ[[rmin1 - 18]] - λ[[rmax1 - 18]]

Sum(θ10[[i, 2]] - avg10)2 (Δλ[[i + rmin1 - rmin]]), i, 1, Length[θ10], 1

flat20 =1

λ[[rmin2 - 18]] - λ[[rmax2 - 18]]


flat30 =1

λ[[rmin3 - 18]] - λ[[rmax3 - 18]]


flat40 =1

λ[[rmin4 - 18]] - λ[[rmax4 - 18]]


flat50 =1

λ[[rmin5 - 18]] - λ[[rmax5 - 18]]


flat60 =1

λ[[rmin6 - 18]] - λ[[rmax6 - 18]]


Flat10 = Abs[avg10 - 2]






4 1_1_Flatness.nb


90

Appendix D|

Mathematica Code for FWHM

Plot

This code was used for the plots in Figs. 4.23 and 4.24. This specific example

is for TRR of the 1:1 sample. The data were plotted and information about the

wavelength and relative transmittance was gathered by right-clicking the plot and

choosing "Get Coordinates". The same process was done for the RRR plot.

91

rmin = 1088;

rmax = 2641;

RRRdata =


1_1 (10_26_2015)\\TRR\\0.csv"];

RRR = RRRdata[[rmin ;; rmax, 3]];

ARRdata =


1_1 (10_26_2015)\\ARR.csv"];

ARR = ARRdata[[rmin ;; rmax, 3]];

Darkdata =


1_1 (10_26_2015)\\Dark_Data_1_1.csv"];

Dark = Darkdata[[rmin ;; rmax, 3]];

λdata =


(10_26_2015)\\ARR.csv"];

λ = λdata[[rmin ;; rmax, 2]];

a =RRR - Dark

ARR;

A = Table[λ[[i]], a[[i]], i, 1, rmax - rmin + 1];

ListPlot[A, PlotRange → All]


92

Appendix E|

Standard Operating Procedure

The following list outlines each deposition procedure:

1. MAKE SURE CURRENT CONTROLLER IS TURNED OFF. Insert tung-

sten boat between electrical contacts and tighten screws.

2. Add the source material to the tungsten boat, packing the material along

the way until material is flush with sides of the boat.

3. Move the first motor to orient the substrate at the correct χv.

4. Go through the following checklist: motor in position, shutter is closed, ma-

terial packed and boat secured, QCM lifetime reads at least 83 %, and correct

deposition program has been selected.

5. Close chamber doors gently as to not knock any material from the boat.

Close the latches on the door.

93

6. Open high-vacuum valve and flip vacuum switch to ‘On’. Push on the cham-

ber door to help create a seal. The high-vacuum pump will automatically

turn on at the correct pressure.

7. Let the chamber sit for at least 3 hours, until the pressure reaches ∼ 2e(−6)

Torr.

8. Turn on the current controller and then press ‘Voltage Out’, then ‘Fine’.

9. Increase the current at ∼ 0.1 A/s.

10. Stop when power reaches about 190 W.

11. Open shutter, start second motor, and zero the thickness.

12. Once the thickness is reached, close shutter and ramp down current. Alter-

natively, if the motor has been programmed to stop automatically, close the

shutter once the second motor has stopped.

13. Let chamber cool for thirty minutes.

14. Switch the high-vacuum to the ‘OFF’ position, close the high-vacuum valve,

turn on the ‘VENT’ switch, and open the nitrogen gas tank.

15. Once the chamber is pressurized, close the nitrogen gas tank and turn off the

‘VENT’ switch.

94

Appendix F|

Non-technical Abstract

Chiral structured thin films (STFs) are arrays of upright parallel nanohelixes that

are fabricated by condensing a collimated vapor from a source material onto a

rotating substrate, with the substrate normal tilted at an angle to the incoming

vapor flux. Instead of continuous rotation, the substrate is rotated in small frac-

tions of one revolution, which allows a significant amount of material to grow at

a time in the direction of the incoming vapor flux. This is referred to as a sub-

deposit. The subdeposits form parallel nanohelixes. This morphology is due to

a self-shadowing process, where the deposited material blocks other parts of the

growing film from the incoming vapor flux, similar to how a tall building in a city

casts a shadow, blocking the sun’s rays from hitting the streets. Another method

involves making two subdeposits 180 apart before rotating back to the position of

the first subdeposit and then introducing the small rotation. This method is called

serial bideposition (SBD), wherein both subdeposits in SBD are of equal height.

95

If the subdeposits are of unequal heights, the deposition technique is known as

asymmetric serial bideposition (ASBD).

Chiral STFs can be used in many applications. Because of the self-shadowing

process described in the previous paragraph, there will be an interconnected space

(void network) between the nanohelixes. This void network can be penetrated by

solutions or gases that change the optical properties of the chiral STF. Thus, chiral

STFs can be used for sensing capabilities.

One can also think of a nanohelix as a spring. Thus, a chiral STF is a bed

of springs. In fact, it has been found chiral STFs possess an elastic displacement

regime, and therefore can react to forces in the linear manner of Hooke’s law.

Before discussing the unique optical properties of chiral STFs, a brief back-

ground in optics is required. In 1865, James C. Maxwell demonstrated that the

phenomenon we call ‘light’ is actually a propagating electromagnetic (EM) wave.

An EM plane wave consists of an electric field oriented perpendicularly to a mag-

netic field, both of which are oscillating and both of which are perpendicular to

the direction of propagation. For most applications, only the characteristics of the

electric field are usually stated.

If the direction of oscillation of the electric field stays the same as the EM wave

propagates, then the EM wave is referred to as linearly polarized. If the direction

of the electric field does not stay in the same plane as the EM wave propagates,

then the EM wave can either be elliptically or circularly polarized, the difference

96

between the two being that the first has an oscillating electric-field amplitude,

and the second has a constant electric-field amplitude. In both cases, the electric

field can rotate clockwise (left-handed) or counter-clockwise (right-handed) with

respect to the propagation direction.

A chiral STF possesses a certain handedness (right or left i.e., counter-clockwise

or clockwise) which can be determined by the sense of the substrate rotation during

growth. Incoming circularly polarized light of the same handedness will get highly

reflected from the chiral STF, while incoming circularly polarized light of the oppo-

site handedness will not. Occurring in a limited portion of the spectral regime, this

circular-polarization-discriminatory phenomenon is called the circular Bragg phe-

nomenon. These characteristics make chiral STFs useful as circular-polarization

filters.

This work focused on the circular Bragg phenomenon exhibited by ASBD chi-

ral STFs. Films were made with various subdeposit-height ratios. Left-handed

and right-handed circularly polarized light were incident upon all the samples for

varying angles of incidence. The spectra of the measured reflectances and trans-

mittances show that ASBD chiral STFs can function as narrow-band circular-

polarization filters. Applications of ASBD chiral STFs may possibly be extended

to resonance cavity mirrors in vertical-cavity-surface-emitting lasers and for sens-

ing purposes with surface multi-plasmonic resonance imaging.

97

Bibliography

[1] M. Faryad and A. Lakhtakia, “The Circular Bragg Phenomenon,” Advancesin Optics and Photonics, Vol. 6, 2014 pp. 225–292.

[2] A. Lakhtakia and R. Messier, Sculptured Thin Films: Nanoengineered Mor-phology and Optics, SPIE Press, Bellingham, Washington, 2005.

[3] J. A. Sherwin and A. Lakhtakia, “Nominal Model for the Optical Repsonseof a Chiral Scultptured Thin Film Infiltrated by and Isotropic Chiral Fluid—Oblique Incidence,” Optics Communications, Vol. 222, 2003 pp. 305–329.

[4] S. Erten, A. Lakhtakia, and G. D. Barber, “Experimental Investigation ofCircular Bragg Phenomenon for Oblique Incidence,” Journal of the OpticalSociety of America A, Vol. 32, 2015 pp. 764–770.

[5] E. Ertekin and A. Lakhtakia, “Optical Interconnects Realizable with Thin-Film Helicoidal Bianisotropic Mediums,” Proceedings of Royal Society (Lon-don) A, Vol. 457, 2001 pp. 817–836.

[6] V. Ross, “The Light-Driven Computer,” Popular Science, Vol. 278, No. 4,2011 pp. 58-59.

[7] M. W. Seto, K. Robbie, D. Vick, M. J. Brett, and L. Kuhn, “MechanicalResponse of Thin Films with Helical Microstructures,” Journal of VacuumScience and Technology B, Vol. 17, 1999 pp. 2172–2177.

[8] A. Lakhtakia, “On Determining Gas Concentration Using Dielectric Thin-Film Helicoidal Bianisotropic Medium Bi-Layers,” Sensors and Actuators B:Chemical, Vol. 17, 1999 pp. 2172–2177.

[9] Y. J. Liu, J. Shi, F. Zhang, H. Liang, J. Xu, A. Lakhtakia, S. J. Fonash, and T.J. Huang, “High-speed Optical Humidity Sensors Based on Chiral SculpturedThin Films,” Sensors and Actuators B: Chemical, Vol. 156, 2011 pp. 593–598.

[10] A. Lakhtakia, M.W. McCall, J. A. Sherwin, Q. h. Wu, and I. J. Hodgkinson,“Sculptured-Thin Film Spectral Holes for Optical Sensing of Fluids,” OpticsCommunications, Vol. 194, 2001 pp. 33–36.

98

[11] S. M. Pursel and M. W. Horn, “Prospects for Nanowire Sculptured Thin-FilmDevices,” Journal of Vacuum and Science Technology B: Microelectronics andNanometer Structures, Vol. 25, 2007 pp. 2611–2615.

[12] S. E. Swiontek and A. Lakhtakia, “Influence of Silver-Nanoparticle Layer in aChiral Sculptured Thin Film for Surface-Multiplasmonic Sensing of Analytesin Aqueous Solution,” Journal of Nanophotonics, Vol. 10, 2016 art. no. 033008.

[13] I. J. Hodgkinson and Q. h. Wu, Birefringent Thin Films and Polarizing Ele-ments, World Scientific, Singapore, 1997.

[14] D. L. Smith, Thin-Film Deposition: Principles and Practice, McGraw–Hill,Boston, Massachusetts, 1995.

[15] M. Faraday, “The Bakerian Lecture: Experimental Relations of Gold (andother metals) to Light,” Philosophical Transactions of the Royal Society ofLondon, Vol. 147, 1857 pp. 145-181.

[16] A. W. Wright, “On the Production of Transparent Metallic Films by the Elec-trical Discharge in Exhausted Tubes,” American Journal of Science, Vol. 13,1877 pp. 49–55.

[17] A. Kundt, “Ueber Doppelbrechung des Lichtes in Metallschichten, welchedurch Zerstäuben einer Kathode herestellt sind,” Annalen der Physik andChemie (Leipzig), Vol. 27, 1886 pp. 59–71.

[18] H. König and G. Helwig, “Über die Struktur schräg aufgedampfter Schichtenund ihr Einfluß auf die Entwicklung submikroskopischer Oberflächen-rauhigkeiten,” Optik, Vol. 6, 1950 pp. 111–124.

[19] B. A. Movchan and A. V. Demchishin, “Study of the Structure and Propertiesof Thick Vacuum Condensates of Nickel, Titanium, Aluminum Oxides, andZirconium Dioxide,” Physics of Metals and Metallography, Vol. 28, 1969 pp.83–90; translated from: Fizika Metallov i Metallovedenie, Vol. 28, 1969, 653–660.

[20] J. A. Thornton, “Structure-Zone Models of Thin Films,” Proceeding of SPIE,Vol. 95, 1988 pp. 95–103.

[21] R. Messier and S. Trolier-McKinstry, “Processing of Ceramic Thin Films,”Encyclopedia of Materials: Science and Technology, Elsevier, Amsterdam, TheNetherlands, 2001.

[22] J. A. Thornton, “High Rate Thick Film Growth,” Annual Review of MaterialScience, Vol. 7, 1977 pp. 239–260.

99

[23] I. Hodgkinson, QH. Wu, and J. Hazel, “Empirical Equations for the PrincipalRefractive Indices and Column Angle of Obliquely Deposited Films of Tanta-lum Oxide, Titanium Oxide, and Zirconium Oxide,” Applied Optics, Vol. 37,1998 pp. 2653–2659.

[24] J. M. Nieuwenhuizen and H. B. Haanstra, “Microfractography of Thin Films,”Philips Technical Review, Vol. 27, 1966 pp. 87–91.

[25] T. Motohiro and Y. Taga, “Thin Film Retardation Plate by Oblique Deposi-tion,” Applied Optics, Vol. 28, 1989 pp. 2466–2482.

[26] N. O. Young and J. Kowal, “Optically Active Fluorite Films,” Nature, Vol. 183,1959 pp. 104–105.

[27] A. Lakhtakia and R. Messier, “The Key to a Thin Film HBM: the Motohiro-Taga Interface,” Proceedings of Chiral ’94: 3rd International Workshop onChiral, Bi-isotropic and Bi-anisotropic Media (F. Mariotte and J.-P. Parneix,eds), Périgueux, France, 1994, pp. 125–130.

[28] K. Robbie, L. J. Friedrich, S. K. Dew, T. Smy, and M. J. Brett, “Fabricationof Thin Films with Highly Porous Microstructures,” Journal of Vacuum andScience Technology A, Vol. 13, No. 3, 1995 pp. 1032–1035.

[29] K. Robbie and M. J. Brett, “Sculptured Thin Films and Glancing Angle De-position: Growth Mechanics and Applications,” Journal of Vacuum Scienceand Technology A, Vol. 15, 1997 pp. 1460–1465.

[30] K. Robbie, M. J. Brett, and A. Lakhtakia, “First Thin Film Realization of aHelicoidal Bianisotropic Medium,” Journal of Vacuum and Science TechnologyA, Vol. 13, 1999 pp. 2991–2993.

[31] I. J. Hodgkinson, A. Lakhtakia, Q. h. Wu, L. De Silva, and M. W. McCall,“Ambichiral, Equichiral and Finely Chiral Layered Structures,” Optical Com-munications, Vol. 239, 2004 pp. 353–358.

[32] I. Hodgkinson, QH. Wu, B. Knight, A. Lakhtakia and K. Robbie, “VacuumDeposition of Chiral Sculptured Thin Films with High Optical Activity,” Ap-plied Optics, Vol. 39, 2000 pp. 642–649.

[33] I. Hodgkinson and QH. Wu, “Serial Bideposition of Anisotropic Thin Filmswith Enhanced Linear Birefringence,” Applied Optics, Vol. 38, 1989 pp. 3621–3625.

[34] <http://www.nobelprize.org/nobel_prizes/physics/laureates/1915/>. (Lastaccessed 2/28/2016).

100

[35] E. Reusch, “Untersuchung über Glimmercombinationen,” Annalen der Physik(Leipzig), Vol. 138, 1869 pp. 628–638.

[36] G. Joly and J. Billard, “Quelques champs électromagnétiques dans les pilesde Reusch I. Les vibrations propres d’une pile de deux lames a biréfringencerectiligne ne sont pas orthogonales,” Journal of Optics (Paris), Vol. 12, 1981pp. 323–329.

[37] G. Joly and J. Billard, “Quelques champs électromagnétiques dans les pilesde Reusch II. Piles éclairées sous l’incidence normale par des ondes monochro-matiques planes et uniformes,” Journal of Optics (Paris), Vol. 13, 1982 pp.227–238.

[38] G. Joly and N. Isaert, “Quelques champs électromagnétiques dans les piles deReusch III. Biréfringence elliptique des vibrations itératives; activité optiquede piles hélicoïdales d’extension finie,” Journal of Optics (Paris), Vol. 16, 1985pp. 203–213.

[39] G. Joly and N. Isaert, “Quelques champs électromagnétiques dans les pilesde Reusch IV. Domaines multiples de réflexion sélective,” Journal of Optics(Paris), Vol. 17, 1986 pp. 211–221.

[40] P. G. De Gennes and J. Prost, The Physics of Liquid Crystals, Second Edition,Oxford University Press Inc., New York, USA, 1993.

[41] A. Lakhtakia, “Spectral Signatures of Axially Excited Slabs of Dielectric Thin-Film Helicoidal Bianisotropic Mediums,” European Physics Journal of AppliedPhysics, Vol. 8, 1999 pp. 129–137.

[42] A. Lakhtakia, “Bragg-Regime Absorption in Axially Excited Slabs of Dielec-tric Thin-Film Helicoidal Bianisotropic Media,” Microwave and Optical Tech-nology Letters, Vol. 22, 1999 pp. 243–247.

[43] V.C. Venugopal and A. Lakhtakia, “On Absorption by Non-Axially ExcitedSlabs of Dielectric Thin-Film Helicoidal Bianisotropic Mediums,” EuropeanPhysical Journal Applied Physics, Vol. 10, 2000 pp. 173–184.

[44] D. M. Mattox, Handbook of Physical Vapor Deposition (PVD) Processing,Second Edition, Elsevier, 2010.

[45] <https://www.alfa.com/en/catalog/013241/>. (Last accessed 12/15/2015).

[46] <http://www.crystran.co.uk/optical-materials/zinc-selenide-znse>. (Last ac-cessed 12/01/2015).

[47] <http://www.lesker.com/newweb/deposition_materials /materialdeposi-tionchart.cfm?pgid=0>. (Last accessed 12/01/2015).

101

[48] <https://www.thorlabs.com/NewGroupPage9.cfm?ObjectGroup_ID=3981>. (Last accessed 12/02/2015)

[49] <http://www.tydexoptics.com/materials1/for_transmission_optics/cvd_znse/>. (Last accessed 12/02/2015)

[50] S. Kasap and P. Capper, Springer Handbook of Electronic and Photonic Ma-terials, Springer, New York, NY, USA 2006.

[51] <https://www.lesker.com/newweb/deposition_materials/depositionmaterials_evaporationmaterials_1.cfm?pgid=zn5>. (Last ac-cessed 12/02/2015).

[52] S. Erten, S. E. Swiontek, C. M. Graham, and A. Lakhtakia, “ExperimentalInvestigation of Circular Bragg Phenomenon Exhibited by a Mirror-backedChiral Sculptured Thin Film,” Journal of Nanophotonics, Vol. 9, 2015 art.no. 090599.

[53] S. E. Swiontek, Supression of Circular Bragg Phenomenon in Chiral Sculp-tured Thin Films Prodcued with Simultaneous Rocking and Rotation of Sub-strate During Serial Bideposition, M. Sc. Thesis, The Pennsylvania State Uni-versity, University Park PA, USA, 2011.

[54] N. J. Severs, “Freeze-Fracture Electron Microscopy,” Nature Protocols, Vol. 2,2007 pp. 547–576.

[55] <http://www.dow.com/assets/attachments/business/gt/infrared_materials/cvd_zinc_selenide/tds/cvd_zinc_selenide.pdf>. (Last accessed 3/1/2016).

[56] Y. Zhu, F. Zhang, G. You, J. Liu, J. D. Zhang, A. Lakhtakia, and J. Xu, “Sta-ble Circularly Polarized Emission from a Vertical-Cavity Surface-EmittingLaser with Chiral Reflector,” Applied Physics Express, Vol. 5, 2012 art. no.032102.

[57] F. Zhang, J. Xu, A. Lakhtakia, T. Zhu, S. Pursel, and M. Horn, “CircularPolarization Emission from an External Cavity Diode Laser,” Applied PhysicsLetters, Vol. 92, 2008 art. no. 111109.

[58] S. M. Pursel, M. W. Horn, and A. Lakhtakia, “Blue-Shifting of Circular BraggPhenomenon by annealing of Chiral Sculptured Thin Films,” Optics Express,Vol. 14, 2006 pp. 8001–8012.

[59] S. M. Pursel, M. W. Horn, and A. Lakhtakia, “Tuning of Sculptured-thin-filmSpectral-hole Filters by Postdeposition Etching,” Optical Engineering, Vol. 46,2007 art. no. 040507.

102

[60] V. C. Venugopal, “Three-dimensional Periodic Chiral Sculptured Thin Films,”Journal of Nanophotonics, Vol. 7, 2013 art. no. 073502.

103

Documents

REFLECTION AND TRANSMISSION OF OBLIQUELY INCIDENT LIGHT BY ASYMMETRIC SERIAL